FIRST DIVISION, INNER HOUSE, COURT OF SESSION
 CSIH 77
OPINION OF THE COURT
delivered by LORD CARLOWAY, the LORD PRESIDENT
in the reclaiming motion
THE SCOTCH WHISKY ASSOCIATION AND OTHERS
Petitioners and Reclaimers;
THE LORD ADVOCATE
THE ADVOCATE GENERAL
Petitioners: O’Neill QC, M Ross; Brodies LLP
First Respondent: Moynihan QC, Irvine; Scottish Government Legal Directorate
Second Respondent: Simpson QC, MacGregor; Office of the Advocate General
21 October 2016
Introduction and Procedure
 Almost five years ago, on 31 October 2011, the Scottish Government introduced a very short Bill to the Scottish Parliament. The Bill was passed, following a Stage 3 debate, on 24 May 2012 with 86 MSPs voting in favour, only 1 against, with 32 abstentions. The Bill became the Alcohol (Minimum Pricing) (Scotland) Act and received Royal Assent on 29 June 2012. The Act amends the Licensing (Scotland) Act 2005 by adding a new paragraph to Schedule 3. This paragraph reads:
“6A(1) Alcohol must not be sold … at a price below its minimum price”.
The minimum price is to be calculated according to a formula, viz.: MPU x S x V x 100 (minimum price per unit x strength of the alcohol x volume of the alcohol in litres).
 The Act empowers the Government to specify the MPU “by order”. On 14 May 2012 the Cabinet Secretary for Health, Wellbeing and Cities Strategy (now the First Minister) announced that 50p per unit was the preferred minimum price. A draft order (The Alcohol (Minimum Price per Unit) (Scotland) Order 2013) was published, setting the MPU at that level. In practical terms, were the Act and Order to come into force, a 70cl bottle of spirits with an ABV (alcohol by volume) of 40% (70°) would retail at a minimum of £14.00 (50 x 40 x 0.7). A 75cl bottle of wine with an ABV of 12% would cost at least £4.50. A pint of lager with an ABV of 4% would cost £1.14 or more.
 In terms of the Act, the minimum pricing provisions are to expire at the end of a 6 year period, unless the Government provide “by order” otherwise after 5 years. There is a requirement that they lay before Parliament a report on the operation and effect of the minimum pricing provisions during the 5 year period. The report must contain information on the effect of minimum pricing on the “licensing objectives” of the 2005 Act. These include the prevention of crime and disorder, protecting and improving public health and protecting children from harm (2005 Act, s 4). It must also advise on its effect on license holders and alcohol producers, both of whom require to be consulted during its preparation. In this sense, the provisions can be regarded as experimental in nature. However, the legislation is not yet in force and, without that, no experiment can take place.
 As will be seen, the Government’s contention is that the Act will bring health benefits of one sort or another to at least part of the population. On 19 July 2012, the petitioners, who represent a variety of alcohol related interests, not least whisky distillers and wine and spirit importers, presented a petition for judicial review challenging the legislation’s lawfulness.
 Originally, one of the grounds of challenge was that the legislation was outwith the competence of the Scottish Parliament, since it related to a reserved matter in terms of the devolution settlement contained in the Scotland Act 1998. This was not insisted upon, following Imperial Tobacco v Lord Advocate 2013 SC (UKSC) 153. A second ground was that the legislation restricted the freedom of trade guaranteed by Articles 4 and 6 of the Union with England Act 1707 and its English counterpart. That was rejected by the Lord Ordinary in the Opinion accompanying his interlocutor of 3 May 2013. This left three grounds, all based upon the legislation’s incompatibility with European Union law, thus rendering it “not law” (1998 Act ss 29(1) and (2)(d); 54(2) and (3) and 57(2)). One was that the legislation was in breach of Article 6(2) of Regulation (EC) 110/2008, which provides that “member states shall not prohibit or restrict the import, sale or consumption of spirit drinks which comply with this Regulation”. It had been argued that this stopped any measure restricting the consumption of spirits. This too was rejected by the Lord Ordinary. None of these grounds were revived in this reclaiming motion (appeal).
 The remaining grounds are, first, that the legislation would breach Article 34 of the Treaty on the Functioning of the European Union, by imposing a quantitative restriction on the import (free movement) of goods. The first respondent accepts that it does have that effect, but maintains that the measure is justified in order to protect health and life under Article 36. Secondly, there is an argument based upon the operation of the Common Organisation of the Markets in agricultural products (then Regulation (EC) No 1234/2007 as amended by No 491/2009; the Common Agricultural Policy). The Lord Ordinary, who refused the prayer of petition in his interlocutor, held that the measure was objectively justified under Article 36. He did not consider that the argument on the CAP was well founded, given the limited effect of the Regulation on prices.
 The reclaiming motion seeks to review the Lord Ordinary’s interlocutor. In the course of the hearing before him, the petitioners had suggested that he might make a Reference for a preliminary ruling to the Court of Justice of the European Union, if he considered that the answer to any question about the interpretation of EU law was not clear. The Lord Ordinary did not consider that the justification provisions of Article 36 required elucidation. He was satisfied with sufficient certainty that the petitioners’ arguments based upon Regulation 1234/2007 fell to be rejected. He did not therefore regard it as appropriate or necessary to refer any questions to the CJEU. However, on 3 July 2014, after a hearing of the reclaiming motion before an Extra Division, the case was referred. The Judgment of the CJEU was obtained on 23 December 2015. Thereafter, the hearing of the reclaiming motion fell to be resumed and determined in light of, inter alia, the CJEU’s answers (infra) to the questions posed in the Reference.
 Looked at from a distance, the litigation throws into sharp focus a measure, which a government deems to be one which will serve to improve the nation’s health, with the commercial interests of the producers and sellers of alcohol, whose ability to trade across the EU is protected by Article 34 and the Common Market. This is not an uncommon dilemma (see World Health Organisation: Global Strategy to reduce the harmful use of alcohol (2010) c 1, para (6(d) “Balancing different interests”).
The History of the Legislation
The Policy Memorandum
 The Government had announced its intention to look at minimum pricing as early as 2007. A strategy discussion paper, namely Changing Scotland’s relationship with alcohol (Scottish Government, 2008), had been published. Extensive consultation had taken place during the unsuccessful promotion of minimum pricing in the earlier Alcohol etc. (Scotland) Bill 2009.
 The new Bill was accompanied by a substantial Policy Memorandum, prepared by the Government, which contained a large number of references to sundry papers and articles. The Bill was described as part of the strategic approach to tackling alcohol consumption more generally. This approach had been set out in Changing Scotland’s Relationship with Alcohol: A Framework for Action (Scottish Government, 2009). The two strategy papers were referred to, but not formally produced, in the reclaiming motion.
 The Memorandum stated, in an overview (para 3), that the Government considered that the objective of the Bill was that minimum pricing would:
“help reduce alcohol consumption … in particular reducing the consumption of alcohol by harmful drinkers, and reduce the impact that alcohol misuse and overconsumption has on public health, crime, public services, productivity, and the economy as a whole.”
 The Memorandum set out (para 6), as part of the background, the stark fact that, since 2000, enough alcohol had been sold in Scotland annually to enable all adults to exceed the sensible male weekly guideline of 21 units on each and every week. In 2010, average sales equated to 22.8 units per person per week; up 11% from 1994. This was being driven (para 20) by off-trade (largely supermarket) sales, which had increased by 52% over the period, compared with a fall of 29% in the on-trade (public houses), and now accounted for two thirds of all sales. Sales were almost a quarter higher than in England & Wales, where they were in decline.
 There was “clear evidence” (para 6) that increased consumption was producing increasing harm. Alcohol related admissions to hospital, at 40,000 per annum in 2010, had more than quadrupled since the early 1980s and related to both genders and all age groups. Alcohol mortality had more than doubled over the same period. It was almost twice that in England & Wales. Over the last 30 years, Scotland has had one of the fastest growing rates of chronic liver disease and cirrhosis in the world. Alcoholic liver disease ranked alongside heart disease, stroke and cancer as a “big killer” (para 9). Life expectancy in some parts of Scotland was well short of that elsewhere and, the Government believed, alcohol played a significant part in this inequality. Alcohol discharges from hospital in the fifth most deprived communities were about 7.5 times higher than in the most affluent fifth. Alcohol mortality was 6 times greater (para 10).
 The Memorandum set out (para 11) the significant social and economic costs of excessive alcohol consumption. The total was estimated at £3.56 billion per annum, including: £866M in lost productivity; £269M in health care; and £727M in the cost of crime. It was said that alcohol misuse was no longer a marginal problem (para 13). At least 50% of men and 39% of women exceeded the weekly and/or daily sensible drinking guidelines. Considerable numbers of children were drinking alcohol. Over a five week period, some 650 children had been treated in hospital for alcohol related problems. Three quarters of prisoners had an alcohol misuse problem, with over one third being alcohol dependent. At least 70% of assaults were alcohol related.
 The harm caused had become a major challenge affecting Scottish society. If it were tackled through minimum pricing, set in the wider strategic context, the benefits would, according to the Government, be clear. Building a healthy and stable relationship with alcohol would be pivotal to realising the Government’s purpose of creating a more successful country (para 15).
 The general policy objective was stated (para 16) to be “to protect and improve public health and attain social benefits by reducing alcohol consumption”. The World Health Organisation (Global strategy to reduce the harmful use of alcohol (2010)) was cited as indicating that, whereas interventions targeted at vulnerable populations could prevent alcohol related harm, policies addressed at the population as a whole could have a protective effect on the vulnerable whilst reducing the overall level of alcohol related problems. Thus, the Government surmised, both population based strategies and those targeting particular groups, such as harmful drinkers, were required. Cheaper alcohol relative to its strength tended to be bought more by harmful drinkers and, in that sense, minimum pricing was a targeted, as well as a population wide, approach.
 The Government had looked (para 17) at “strong evidence” from studies in Europe, the United States and the Commonwealth to the effect that levels of alcohol consumption were closely linked to price. A systematic review of over 100 studies (Wagenaar et al: Effects of beverage alcohol taxes and prices on consumption: a systematic review and meta-analysis of 1003 estimates from 112 studies (2009)) had found that there was a consistent relationship between price and consumption. The RAND Europe report (Rabinovich et al: The affordability of alcoholic beverages in the European Union: understanding the link between alcohol affordability, consumption and harms (2009)) supported the existence of that link and the idea that alcohol pricing policies were effective as a measure to curb hazardous and harmful drinking. A further study (Booth et al: Independent Review of the Effects of Alcohol Pricing and Promotion: Part A: Systematic Reviews (2008)), commissioned by the United Kingdom Government from the School of Health and Related Research at Sheffield University, found strong and consistent evidence to suggest that price increases had a significant effect in reducing demand.
 The impact of different prices in the off and on-trade was significant (para 21). In 2010, the average on-trade price per unit was £1.34. In the off-trade it was only 45p. Some 11% of off-trade sales retailed at below 30p per unit, 45% below 40p and almost 75% under 50p. Although the affordability of on-trade sales had increased only slowly since 1987, off-trade sales were now about 130% more affordable for beer and 98% for wine and spirits over that period. As that affordability had increased, so had sales risen at the expense of the on-trade market.
 The research at Sheffield University, conducted respectively for the UK and Scottish Governments (Brennan et al: Independent Review of the Effects of Alcohol Pricing and Promotion: Part B: Modelling the Potential Impact [etc] … in England (2008) and Model-Based Appraisal of Alcohol Minimum Pricing and Off-Licensed Trade Discount Bans in Scotland using the Sheffield Alcohol Policy Model (v2): An Update [etc] (2010)), estimated that policies which increased the price of alcohol could bring significant health and social benefits and lead to considerable savings in the health service, criminal justice, and the workplace (para 24).
 Detailed findings were contained in a Business and Regulatory Impact Assessment for Minimum Price per Unit of Alcohol (BRIA) prepared for the Bill (infra). The Government noted (Policy Memorandum, (para 24) that, according to these findings, the model demonstrated a strong link between price increases, reduced consumption and subsequent reductions in chronic and acute health harms, including cancers, stroke, accidents, injuries and violence. Minimum pricing targeted cheap alcohol, which tended to be bought more by harmful than moderate drinkers. It targeted the drinkers who were causing most harm to themselves and society. Cheaper alcohol was also attractive to young people. Moderate drinkers would only be marginally affected. The economy was likely to benefit in terms of a reduction in sick days for all categories of drinkers (moderate, hazardous and harmful) and in unemployment among harmful drinkers.
 The alternative of increased taxation was considered (Policy Memorandum para 28). The Government did not consider that taxation, on its own, would provide an adequate alternative. The Scottish Government had no power to increase taxation on alcohol. The current tax arrangements were inequitable, with different types of alcohol being taxed at different levels. Tax was not linked to ABV. Any altered arrangements could not provide a basis for a policy which was fair to all alcohol producers and sellers. A minimum price based on tax increases could have a disproportionate effect on some products but not others. A taxation regime based on ABV would not comply with EU law.
 Increases in taxation would not necessarily result in a proportionate, or any, increase in the retail price. They were not always reflected in that price. Many supermarkets sold alcohol at below cost and absorbed the tax increases, or offset them against other products. Tax increases did not have a targeted effect on those most at risk of alcohol related harm. This was because harmful and hazardous drinkers consumed a disproportionate amount of cheaper alcohol. Tax increases would have a proportionately greater impact on moderate drinkers. A pricing measure which predominantly affected the off-trade was likely to be more effective at tackling alcohol harms. A tax system which targeted lower cost alcohol would be difficult to administer.
 The Bill was accompanied by Explanatory Notes, prepared by the Government, which echoed much of what was contained in the Policy Memorandum. They are not therefore repeated at length. The Notes did emphasise the increase in alcohol consumption and resultant harm over recent decades. They pointed (para 28) to the Sheffield University research (supra) and, in relation to the setting of the minimum price, detailed the modelling done for prices set at a minimum of 25p through to 70p, at 5p increments. These estimated a reduction in consumption ranging from, for example, 4.3% for a minimum unit price of 45p, 6.7% for 50p and 9.5% for 55p. A 70p minimum price would reduce consumption by 18.4%. With the off-trade discount ban, which had been introduced by the Alcohol (Scotland) Act 2010, these figures would increase to 6.7%, 8.7%, 11.2% and 19.5%.
 The Notes repeated the critical point that the Sheffield modelling demonstrated that minimum pricing would have the greatest impact on the consumption of cheap alcohol, which was mostly drunk by harmful (in excess of 50 or 35 units per week for respectively men and women) and hazardous (21-50, 14-35) drinkers. This was illustrated by a table (para 37, Table 3) which showed the reduction in annual alcohol consumption and the increase in annual spending, for the same minimum prices selected above, for different categories of drinker, as follows:
The reduction of 7.9% for harmful drinkers, if there were a minimum price of 45p, would represent a weekly reduction of 5.7 units, whereas the 2% for moderate drinkers represented a decrease of only 0.1 units.
 The Notes contained tables illustrating health and other benefits. They then addressed (para 47 et seq) the effect on businesses. There would be benefits to the “alcohol industry” as a whole. The modelling, however, had not involved a consideration of which sector (retailers, wholesalers and producers) would benefit or to what extent. A table (para 49, Table 7) covering the estimated effects on the on and off-trades, again using the same figures, produced the following benefits in financial terms:
 The Notes made reference (para 52) to questions in relation to minimum pricing which had been posed to groups representing retailers and producers during the unsuccessful passage of the previous Bill. It was not thought there would be a cross border dimension, as there was in Ireland. The majority of consumers would not purchase alcohol using the internet.
 The final Business and Regulatory Impact Assessment for Minimum Price per Unit of Alcohol contained more detailed material of a similar nature to that which found its way into the Policy Memorandum. It summarised the problem as follows (para 1):
“By global standards, Scotland consumes very high levels of alcohol. Even within the European Union, our consumption is above average and our pattern of consumption, allied to other health factors, means there is a very significant impact which undermines the country’s potential as individuals, families and communities.”
The Government’s strategy (supra) had set out some 40 measures aimed at addressing alcohol related harm, along similar lines to the WHO’s Global strategy (supra). The BRIA set out the various measures which had already been put in place, including: record investment in schemes designed to tackle alcohol misuse, improving prevention and treatment services; a national programme of screening; the reform of local alcohol and drug delivery arrangements; the ban on quantity discounts and alcohol promotions in off-sales; the establishment of a Youth Commission on Alcohol; refreshed advice for parents and carers; and continued work with “industry partners” on joint initiatives to promote responsible drinking, such as Alcohol Awareness Week. The measures, as part of the overall strategy, were underpinned by wider policy initiatives across health, education, justice and the economy. These sought to address the underlying causes of poor health and social disadvantage. The summary continued (para 6):
“Despite these actions, and an economic downturn in recent years, Scotland’s consumption and harm remain at unacceptably high levels. The key component missing from Scotland’s alcohol strategy has been an intervention to address the low price of alcohol”.
 The minimum pricing policy aimed to reduce alcohol consumption and targeted a reduction in consumption of alcohol which was cheap relative to its strength. It achieved this (para 10):
“because it is both a whole population approach and a targeted intervention – it applies to the whole population, but hazardous and harmful drinkers are likely to be affected more than moderate drinkers, in terms of the amount they drink, how much they spend and how much they benefit from reductions in harm”.
 The BRIA stated (para 4.3) that the objective of protecting and improving public health would not be achieved through taxation because: broad tax increases did not have a targeted effect on those at risk (ie hazardous and harmful drinkers); straightforward increases in tax would impact on high price products as well as cheap ones and thus have a proportionately greater effect on moderate drinkers; tax increases did not necessarily result in a proportionate, or any, rise in prices because increases were not always reflected in the price (10 grocery retailers sold at below cost to varying extents); an increase based on price would disproportionately affect the market; and taxation based on units of alcohol would not be compatible with the current EU regime. Finally:
“Even if it was possible to formulate a scheme of taxation proportionate to the number of units … it would remain possible for retailers (particularly large supermarket chains) to absorb this tax by loss leading on alcohol and putting up the price of other products … Absorption could not be prevented by prohibiting sales below cost plus tax because cost would always be susceptible to manipulation.”
This was because the cost was subject to variation, manipulation and cross-subsidisation. The declared cost may bear little relation to the actual cost. If taxation were set at a level similar to the minimum price, the rate across all alcohol would be considerably higher than that currently in place. Minimum pricing had the advantage over taxation because it created certainty. It was not open to absorption. It did not encourage cross-subsidisation between products. It was easier to understand, measure and enforce. Moderate drinkers would be largely, or completely, unaffected. It targeted those who would benefit most.
 The BRIA echoed (paras 4.6/7) the “sophisticated econometric modelling” undertaken by Sheffield University, which estimated that a 50p MPU would lead to: an overall fall in consumption of 5.7%; 60 fewer annual deaths in the first year, rising to over 300 (a fall of 17%) in the tenth year; 1,600 fewer hospital admissions in the first year, rising to 6,500 in the tenth (10% decrease); 3,500 (2%) fewer crimes per year; 32,000 fewer days off work and a reduction in unemployment of 1,300. The estimated total harm reduced was valued at £942M by year 10.
 The BRIA contained a section (5) on the effects on different groups, including consumers, producers, distributors and retailers. The extent of any impact on producers would depend on the quantity of alcohol which they continued to produce at below the minimum pricing level. Some 70% of off-sales came from the large supermarkets (para 5.39) who had substantial buying power and the ability to negotiate lower prices from suppliers and producers. The views of the petitioners were noted (para 5.40). Given that some retailers sold alcohol at below cost as a competitive strategy, moderate drinkers were subsidising heavy drinkers (para 5.42). The retailers might reduce the prices of goods, which were currently subsidising low prices on alcohol such as DVDs, non-alcoholic drinks and health and beauty products.
 There was “no consistent view from the industry” on the effect on producers (para 5.49). The petitioners thought that any increased revenue would be kept by the retailers. Some whisky producers stated that retailers would not discuss pricing policy with their suppliers. The supply side reaction was not known (para 5.101) and there were differing views. It was difficult to predict the effect on producers. There was “no consensus from industry” (para 5.114) on what would happen to pricing and hence the effect on the market. Some considered that all prices would be effected, but others disagreed. Some thought own label brands would be “decimated”. There was “no clear consensus” (para 5.115) on whether supermarkets would continue to sell their “own label” spirits (26% of the off-trade market). One company said (para 5.118) that it was not possible to make any forecast before seeing what the shift in consumer behaviour would be.
 It was expressly stated that the policy was “innovative and largely untested” (summary, para 28), albeit that it had been based on “a wealth of international evidence on the relationship between price, consumption and harm”; hence the review provisions.
Other EU States’ Opinions
 On 25 June 2012 the Scottish Government notified the draft Order to the EU Commission in terms of Article 8 of the Technical Standards Directive (98/34/EC). There was considerable interest from other Member States. Bulgaria observed that their wines sold in Britain for about £3.00 per bottle. Minimum pricing would create many obstacles to trade and the export of wines to Britain would be significantly affected. It would lead to a change in consumption trends and probably to a contraction in the market. Despite that view, it was said that the measure would not have an effect on health. Minimum pricing could give rise to reciprocal measures outwith the EU and effect EU exports accordingly.
 Spain objected to what it regarded as a measure designed to eliminate cheaper products from the market. France argued that tax increases or local measures would be less restrictive. Although Italy was more receptive of the health arguments, it said that the social effectiveness of the measure was not clear. Moderate drinkers could be disproportionately affected. The narrowing of prices between low and medium quality brands could result in the confusion of the consumer. The setting of a minimum price for health reasons would be a dangerous precedent. Those countries with the highest prices for alcohol were those where the greatest alcohol related health problems were to be found.
 Portugal considered that the measure would have a dramatic effect on the market, particularly on its exports to the UK. It would raise the price of Portuguese wine, which sold below the minimum price, by between 10 and 60%. As with Spain, Portugal maintained that a dangerous precedent would be set with regard to third countries. There was:
“no causal link between this measure and efficiency while protecting consumer health. In fact, there is nothing to indicate that the introduction of a minimum unit price … would result in decreased alcohol consumption”.
Inappropriate consumption was declining in Scotland and there were more efficient ways of raising alcohol awareness.
 Denmark’s response was relatively short and legalistic. It called for greater specification. The Netherland’s view was also short. It was that the principle of free movement of goods was not an “a priori hindrance” to the proposed measures, but the case law of the CJEU had to be taken into account. Germany’s approach was not dissimilar.
 Poland was more forthright and expansive. It did not consider it likely that the measure would be effective. It would increase “social and economic poverty”. It was “not effectively directed at the low-income minority who drink to excess and irresponsibly”. Poland stressed the barrier that would be created for new products. The Government had exaggerated Scotland’s problem relative to that of other countries. The measure was disproportionate. It would penalise those who drank responsibly. It would hit all consumers. Some of the contentions in the BRIA came under challenge. It would not be difficult for supermarkets to evade the law by establishing altered supply models and encourage customers to purchase via the internet. More effective and less restrictive measures were available and apparently working. Romania opposed the measure on the view that the CJEU had rejected health objectives as a justification for minimum pricing where alternatives were available. The measure would penalise responsible consumers for the behaviour of a minority.
 Austria, whilst pointing out that its minimum pricing for cigarettes had fallen foul of EU law, thought that it would be welcome from the health perspective, even if it was doubtful whether it would solve the alcohol problems.
 Ireland supported the measure. In 2012, its National Substance Misuse Strategy Steering Group had published a report which elaborated on the enormous harm that the misuse of alcohol did to Irish society. These were summarised in similar terms to the BRIA (supra). More than 1.5 million Irish people drank in a harmful pattern. The Steering Group had recommended increasing the price of alcohol and, in addition, introducing a legislative basis for minimum pricing. Ireland’s response continued:
“The rationale for the latter is clear. It is to target hazardous drinkers and adolescents who are sensitive to price changes of alcohol, simply because minimum pricing – as opposed to fiscal measures – is aimed at preventing the sale of alcohol at very cheap prices. A minimum pricing regime is thus a proportional policy exigency that allows the state to engage another parameter to deal with managing the supply of alcohol for the purpose of preventing its misuse.”
Ireland considered that the policy was aimed at harmful drinkers and would only have a marginal effect on moderate drinkers. Hazardous and harmful drinkers drank proportionately more cheaper alcohol. The measure targeted cheaper alcohol because the price was determined by the amount of pure alcohol in the product. Minimum pricing would lead to reductions in health, crime and employment harms. Ireland’s “strong view” was that minimum pricing was a proportionate measure under Article 36.
The EU Commission Opinion
 The Commission itself produced an Opinion on 26 September 2012. The Opinion had regard to the “Dassonville formula”, whereby all trading rules capable of hindering, directly or indirectly, actually or potentially, intra EU trade were to be considered as having an effect equivalent to quantitative restrictions (C-8/74 Procureur du Roi v Dassonville  ECR 837 at para 5). However, “selling arrangements” did not have this effect so long as they affected the marketing of domestic products and those from other EU states “in the same manner” (C-267/91 & C-268/91 Keck and Mithouard  ECR I-6097 at para 16).
 In short, the Opinion was that minimum pricing would not breach Article 34 unless it discriminated against imports. However, minimum pricing did not take into account imports which had lower production costs compared to equivalents in the domestic market. A minimum price could cancel out the competitive advantage of an importer whose costs were lower than those of the domestic seller. Thus, the Opinion reasoned, French brandy, which retailed at less than 50p per unit, would be discriminated against in its competitive relationship with Malt whisky, which was seldom sold at that low price level. New entrants into the UK cider market would be similarly effected, if they wished to sell at a “discount” in order to launch their products. Accordingly, minimum pricing, although applicable to domestic and imported products without distinction, was capable of having an adverse effect on imports and was thus a quantitative restriction in so far as it prevented the lower cost of the import being reflected in the retail price.
 The Opinion went on to consider the justification and proportionality of the measure in terms of Article 36. The measure had to be shown to be necessary in order to achieve its objective. It had to be proportionate to this aim; that is the objective must not be capable of being achieved by any other means less restrictive of intra-EU trade. The aims of the Bill to reduce overall alcohol consumption, and in particular cheap alcohol sales through the off-trade, were fully recognised. The Opinion continued:
“Keeping in mind that most of the studies prove and there is a general agreement that affordability does have effect of (sic, ?on) drinking patterns, the question is only about the best way to exploit this tendency. Hence, the Commission does not disagree with the proposition that increases in the prices of alcoholic drinks, could, other things being equal, be expected to lead to reduced demand for those drinks. The focus in this opinion is rather on whether a MUP policy, which would lead to higher prices of many alcoholic drinks and hence to an expectation of reduced consumption, is likely to be the least market distorting policy that could be introduced to produce such an outcome.”
If the goal were to reduce alcohol consumption by increasing prices, that could be achieved by increasing tax “across the board”. This would not cause the market distortion that would be likely to result from minimum pricing (cf C-216/98 Commission v Greece  ECR 1-8921 at para 31). It “would be a better option to address the alcohol consumption problem without having the adverse effects as it would impact on all products equally”. Minimum pricing would encourage retailers, and supermarkets in particular, to sell more alcohol because of the higher margins on products affected by the policy.
 Specific attention was paid to the Government’s arguments. The Commission was not persuaded that any rise in prices would not be passed on to the consumer. It has been said in Commission v Greece (supra, at para 32) that:
“The ability of manufacturers and importers not to pass on increases in excise duty on their products is in any event limited by the extent of their profit margin, with the result that excise duty increases are sooner or later incorporated in retail selling prices.”
The existing EU Directives allowed states a degree of discretion in formulating fiscal policies. The BRIA modelling did not distinguish between different types or strengths of product within a category. The Government could adopt measures specifically related to the particular problem areas in Scotland. These would be more likely to be effective than measures aimed at the whole population. The Commission concluded that the measure “may” create obstacles to the free movement of goods, contrary to Article 34, and appeared to be disproportionate under Article 36. It ought not to be adopted without the Government giving due consideration to the Commission’s “remarks”.
 The lead committee for the Bill was that of Health and Sport. This was the same Committee as had considered the previous Bill in February 2010, to which some reference was made in the submission in the reclaiming motion (infra, evidence of Dr Meier of Sheffield University). Stage 1 of the Bill commenced on 10, and continued on 17 and 24, January 2012. The Committee received written responses from, and heard the views of, a variety of professionals involved in the assessment of the effects of alcohol on the individual and the community in terms of both health and economics. In due course, the Lord Ordinary was to found upon some of the evidence given at the Committee stage. What follows is selected from the passages cited by him.
 First, there was the evidence of Anne Ludbrook, professor of health economics at Aberdeen University, at the session on 10 January. She said that minimum pricing would have some effect across all income groups, but it would have the most effect on those purchasing the most alcohol, where the impact was desirable. There were difficulties with other pricing interventions. Taxation was not necessarily passed on into prices. It affected all products, whatever price they were at the moment. The Sheffield University model showed that minimum pricing would be more effective than an across the board tax increase. It showed much higher reactions to minimum pricing among the heavier drinkers, because there was a tendency for them to trade down to lower cost products when there was such a tax increase. Evidence from Sweden had shown that, if price increases were targeted at the lowest cost products, there would be a bigger impact on consumption for the same average increase.
 Secondly, Peter Rice, chair of the Royal College of Physicians and consultant addictions psychiatrist in Tayside, added that he did not think it was correct, as one of the MSPs had put it, that better-off people did not notice prices. They did and responded to them. They did trade down. There was a level of speculation about the response of the industry to minimum pricing. Some brands might increase their prices to maintain their differential with cheaper products. The behaviour of the trade was a “big variable”. Prices in supermarkets had gone up in the last few years, but “we do not know why – they do not give us that information, as it is commercially confidential”.
 Evelyn Gillan, chief executive of Alcohol Focus Scotland, said that her organisation had never regarded minimum pricing and taxation as alternatives. They could be complementary. However, the difficulty was that raising taxation would not necessarily increase the price of the very cheapest products to the level at which consumption would reduce. Taxation policies had in the past been regarded as the most effective way of increasing price, but minimum pricing had come to the fore in recent years because people had seen that tax increases were not always passed onto the consumer. For example, in previous years some big supermarkets had advertised “tax-busting” prices; saying that the retailer would be absorbing the duty increase. If taxation was to be used on its own, it would have to be increased significantly in order to bring some of the cheaper products up to the required price.
 In the evening session, Timothy Stockwell, professor at the University of Victoria BC, made a presentation on the effectiveness of minimum pricing in Canada. Asking himself the question of “why introduce minimum pricing”, when across the board tax increases would reduce average consumption, Prof Stockwell said that it was known that the heaviest drinkers gravitated towards the cheapest alcohol. Young people and high risk drinkers were especially responsive to minimum pricing. Prof Stockwell was effusive in his praise of the proposed legislation, stating that the Bill would “without a shred of doubt” save lives, reduce healthcare costs, prevent death and injury on the roads, prevent birth defects, and reduce public violence and a range of other things. He agreed with the likely impact as modelled by Sheffield University.
 At the session on 24 January, Jonathan Chick, professor at Queen Margaret University, explained that most of the previous literature had been about across the board tax changes. Minimum pricing was not an example of that, but was to do with eliminating very inexpensive alcohol. When taxation rose, heavy drinkers traded down. Minimum pricing prevented trading down below the minimum. That is why he saw it as an approach that would affect heavy drinkers. He later added that it was a more effective method than taxation for reducing consumption in the harmful drinking group. Minimum pricing meant that they could no longer buy alcohol cheaply. That would help them and their families.
 Meantime, John Holmes, public health research fellow at the University of Sheffield, had asked himself the key question of whether the policy would disproportionately hit moderate drinkers on low incomes, compared with moderate drinkers on higher incomes, and whether it would have an impact on harmful drinkers on high incomes. His research had shown that harmful drinkers in all income groups bought significant proportions of cheap alcohol. They focused their spending on the off-trade. More than half of their off-trade spending was at less than 50p per unit. That applied to all income groups. In all groups, harmful drinkers spent less per unit than moderate drinkers. Thus the policy would have a bigger effect on harmful drinkers than on moderate drinkers.
 The Stage 1 debate in Parliament took place on 14 March 2012 and the final Stage 3 debate on 24 May 2012. As already noted, there was only one MSP who opposed the Bill’s enactment.
The Lord Ordinary’s Approach (2013 SLT 776;  3 CMLR 34)
 It is important to observe in limine that, although the submissions (infra) made very little mention of the Lord Ordinary’s reasoning, this is an appellate process and not one at first instance. The assessment of fact is primarily for the Lord Ordinary to resolve upon the material presented to him. This appellate court should not ignore his assessment and engage in an entirely new investigation, regardless of his views, unless he could be shown to have erred in his approach to the task. The import of this will be considered in due course, having regard, inter alia, to the parties’ respective applications to have new material considered.
 The petitioners had argued primarily that the clear and constant jurisprudence of the CJEU had been that minimum pricing was always unlawful in terms of Article 34 and could never be justified under Article 36. The Lord Ordinary had regard to the Commission’s Opinion (supra), although he described it, and the view of the member states (supra), as “of interest; but no more than that” (para ). Ultimately he was not persuaded (para ) that the authorities relied upon by the petitioners supported their central proposition. He observed (para ) that C-231/83 Cullet v Centre Leclerc  ECR 305 was to the opposite effect. Had the minimum prices for petrol in that case inevitably been an obstacle to freedom of movement under Article 34, the CJEU would not have needed to consider, as it did, whether they were justified under Article 36.
 The Lord Ordinary noted (para ) that the CJEU’s role in reviewing national measures was a supervisory one. The crucial question was whether there was an objective justification for the measures under review. The fact that there may be controversy, and that the evidence may not all be one way, did not preclude a conclusion that the necessary justification had been made out. The CJEU afforded a measure of respect (margin of appreciation) to the national authorities (C-110/05 Commission v Italy  2 CMLR 34, at para 65; C-434/04 Ahokainen and Leppick  ECR I-9171, at paras 32 and 39) and the national courts should do the same.
 The Lord Ordinary founded (para ) upon dicta in Sinclair Collis v Lord Advocate 2013 SC 221 (at para , following Commission v Italy (supra at paras 59-67)) that a measure requiring justification in terms of Article 36 had to be appropriate for securing the public interest objective and should not go beyond that which was necessary to attain it. States could determine the degree of protection and the way in which it was to be achieved. They had to be allowed a margin of appreciation in that regard. They did not require to prove positively that no other conceivable measure could attain the objective. They were entitled to attain it by introducing general and simple rules, which would be easily understood and applied and easily managed and supervised. The state had to demonstrate that the measure was objectively proportionate; that is appropriate and necessary. Demonstrating, by reference to extraneous reasoning, that the measure was capable of attaining its stated purpose did not involve an examination of the subjective reasoning of the decision making body, but a determination of whether, as a matter of objective fact, there existed material justifying the measure in terms of physical health.
 The Lord Ordinary found (para ) that the aims of the legislation were “the reduction of alcohol consumption, … in particular … by hazardous and harmful drinkers”. These aims were “directed principally towards the protection of health and life, though other consequential (largely public order and economic) benefits are also anticipated”. The measures were intended “to strike at alcohol misuse and overconsumption” (para ) and “to get people to build a healthy and sensible relationship with alcohol”. The selection of 50p had been in an attempt to achieve a reasonable balance between public health and social benefits on the one hand and intervention in the market on the other. The major problem was the “excessive consumption of cheap alcohol”. The legislation sought “to address this by increasing the price”. The reduction of alcohol consumption generally, and by hazardous and harmful drinkers in particular, were legitimate aims in terms of Article 36.
 In relation to the challenge that the legislation was not suitable to reduce consumption by hazardous and harmful drinkers, the Lord Ordinary found that (para ):
“Whilst it is correct that there are a greater number of hazardous and harmful drinkers in higher income than in lower income groups, the harmful drinkers in the lowest income quintile consume far more alcohol per head, and are the source of much greater health related and other harm, than harmful drinkers in the higher income quintiles. There is also clear evidence that the greatest alcohol-related harm is experienced by those who live in the most deprived areas.”
Accordingly, the Lord Ordinary reasoned, there was objective evidence that the measures were appropriate to achieve their aims. That left (para ) the “critical issue” of the effectiveness of minimum pricing and alternative measures, notably taxation increases. If the alternatives “would not be just as effective as minimum pricing in achieving the legitimate aims, minimum pricing would be necessary and proportionate”.
 The Lord Ordinary spent some time examining this issue. He considered (para ) that there was force in that passage of the BRIA which observed that minimum pricing had the advantage of certainty. It was not open to absorption. It did not encourage cross-subsidisation. It was easier to understand, measure and enforce. The EU regime for the imposition of duty did not easily permit increases in taxation relative to ABV. A system, which resulted in higher prices for higher alcohol strength products, was more consistent with the objective than one resulting in similar prices for alcohol of different strengths. Under minimum pricing, alcohol, which was cheap relative to strength, could be targeted. The EU regime precluded taxation being used in this way.
 If prices were increased across the board, consumers of high priced alcohol could choose to drink less or switch to cheaper products. This option did not apply to those already drinking the cheapest alcohol. Total sales were reduced more when price rises were concentrated on the latter, because consumers of the cheapest products did not have the option of switching down and heavier drinkers tended to buy disproportionate quantities of cheap alcohol. If tax were increased to raise the price of alcohol that was too cheap, the same rates would have to be applied to more expensive alcohol. “Targeting – in the sense of focusing duty increases on cheaper alcohol – is not possible” (para ).
 The taxation route favoured by the petitioners and the Commission would involve greater price rises, and would affect more products, than minimum pricing. The argument that, since this would decrease consumption, the aims would be better served using this method presupposed that the aim was to reduce consumption to the maximum extent possible, regardless of the possible economic and social consequences. That was not the aim.
 The Lord Ordinary focused (para ) on the protection of health and life as ranking first among the interests protected by Article 36. It was for the state to decide, within the limits of the Treaty, what degree of protection was required (Commission v Italy (supra) at paras 61 and 65; Sinclair Collis v Lord Advocate (supra) at para 54). Inherent in the legislation was a judgment on the level of protection. The best way of maximising reduction was to focus price increases on cheaper alcohol. Although socio-economic considerations, such as concerns for the on-trade and moderate drinkers, could not be used to justify measures which were more restrictive than the alternatives, they were matters which could be taken into account in deciding the level of protection. Even if taxation could achieve price increases equivalent to minimum pricing, the outcome would be materially different; notably an increase in the prices of many products. Moderate drinkers and the on-trade would be more affected. The level of protection would be greater than that thought to be appropriate and proportionate. For all of these reasons, the Lord Ordinary concluded (para ) that there was objective justification for the view that the alternatives would be less effective in achieving the legislative aims.
 In relation to the Common Agricultural Policy, the contention of the petitioners was that minimum pricing undermined The Common Organisation of the Market in relation to wine, other fermented beverages, and agriculturally produced ethyl alcohol brought about by Regulation (EC) 1234/2007. Article 113c (as inserted by Regulation (EC) 491/2009) permitted Member States to lay down marketing rules to regulate the supply of wines, but these required to be proportionate and (113c(1)(b)) could not “allow for price fixing, including where prices are set for guidance or recommendation”. Article 128 prohibited customs duty on “trade with third countries” and any quantitative measures having the equivalent effect. The legislation breached both articles.
 The Lord Ordinary noted (para ) that, where a matter had been the subject of exhaustive EU harmonisation, any national measure had to be assessed under reference to the harmonising provisions rather than the Treaty. States were able to act on matters, which were not the subject of specific harmonisation, provided that the measure did not undermine, or create exceptions to, the Common Organisation of the Market. Regulation 1234/2007 did not regulate price or deal with the protection of health. It was not a fully harmonised measure dealing comprehensively and exhaustively with the designated products. National measures could deal with matters not covered by the Regulation. There was no conflict between the Regulation and minimum pricing. Article 113c related to the period up to first marketing by the producer. It was aimed at prohibiting price fixing between producers. It did not prohibit or restrict the ability of a state to provide for minimum pricing at the retail stage on the grounds of the protection of health. Although Article 128(b) was similar to Article 34, and there was no equivalent to Article 36 in the Regulations, that did not support the conclusion that the Regulation made exhaustive provision for the Common Organisation of the Market in each of the sectors specifically dealt with (para ).
The Reference to the CJEU
 Having introduced the general area for consideration, the Reference described the “affected market”, according to Government materials available at the stage of the enactment of the Bill. Prior to the Reference, both parties had been allowed to lodge a considerable amount of new information which had not been before the Lord Ordinary. By interlocutor dated 11 October 2013, the court had appointed parties to be heard on 28 and 29 November, not only in relation to whether there ought to be a Reference to the CJEU but on “the evidential basis on which the reclaiming motion should proceed” (ie what material could be looked at). Although there is no interlocutor expressly dealing with that issue, that of 29 November 2013 appointed:
“each party to lodge no later than 6 December 2013, for inclusion in the appendix to the reclaiming print any supplementary materials upon which they wish to rely in support of their contentions on the disputed factual issues.”
This gave parties carte blanche at least to lodge new material. The significance of this will be addressed in due course (infra), but it is sufficient to observe at this stage that the issue of what material could be looked at was not decided, but rather formed the basis of a question for the CJEU.
 Amongst the new material was the NHS Health Scotland survey of 2013 (Monitoring and Evaluating Scotland’s Alcohol Strategy) (MESAS). This stated that, in 2011, 60% of off-trade sales was at below 50p per unit. Although the Reference stated that the materials available at the time of the consideration of the Bill did not cover the effect of the measure on intra EU trade, a further paper from NHS Scotland in 2013 (The price distribution of wine from different countries of origin sold in Scotland’s Off-Trade) (August 2013) gave percentages of the wine from EU countries selling at less than 50p per unit. Those with the highest percentage were Bulgaria (91%, being 0.3% of the market), Romania (86%, 0.1%) and Portugal (72%, 0.7%). They were followed by Italy (63%, 15%), Spain (62%, 8%), France (44%, 10%) and Germany (33%, 1.2%). Significant quantities of imported beer would also retail at below 50p per unit. Cider generally came from England and most of it sold at below 50p per unit. As already described, each of these countries had given its own estimate of the impact which the measure would have on their exports to the United Kingdom. Those with the highest percentage of cheap wine had all written to the Commission with their views.
 The Reference cited a report, which had been before the Lord Ordinary, commissioned by the petitioners from two economists (Yarrow and Decker: Economic analysis of the impact of minimum pricing on alcoholic beverages in Scotland (2012)). It stated that wine producers, whose costs were sufficiently low, would lose a competitive advantage if minimum pricing was to apply to their products. The effect of eliminating that advantage would benefit the retailer, whose profit margin would be increased, and economically disadvantage the consumer. Some retailers might spend less effort promoting products otherwise retailing at just above the minimum price. This might adversely affect “medium cost producers”.
 The Reference mentioned parts of the Memorandum and the BRIA, both of which had accompanied the Bill. It said that the case for minimum pricing, as presented to the Parliament, had been based on the Sheffield University modelling. Later work by the University, conducted in England & Wales on the basis of a 45p minimum price, had demonstrated that, of those in the lowest income quintile of the population: 4.8% were “harmful”(supra) drinkers; 10.59% were “hazardous” (supra); 56.5% were “moderate” drinkers; and 27.8% were abstainers. The figures for the highest income quintile were 7.7%, 25.7%, 59.8% and 6.8% and for the second highest they were 4.9%, 20.3% and 65.8% for harmful, hazardous and moderate drinkers. Harmful drinkers within the lowest quintile bought 30.8 units of cheap alcohol per week. The number of units bought declined with affluence to 13.6 units for the highest quintile. The hazardous drinker in the lowest quintile bought 7.8 units whereas in the highest quintile he bought only 2.7 units. The modelling indicated that the lowest income quintile would drink 300 units less per annum, whereas the figure for the highest would be 34 units. The equivalent for hazardous drinkers would be 42 units for the lowest quintile.
 The petitioners’ contention on these statistics was that hazardous drinkers did not significantly rely on cheap alcohol. The Reference was critical of the BRIA’s reliance on Wagenaar (supra) to demonstrate elasticity, having regard to another post enactment paper (Nelson: Does heavy drinking … respond to higher alcohol prices and taxes (2013)), lodged after the Lord Ordinary’s decision, which was said to cast doubt upon the proposition that policies relying exclusively on price control or higher taxes were effective as a means of reducing “abusive” drinking. The Reference was also critical of the modelling as descriptive of the potential health and societal harm because of the absence of input covering different incomes and the price elasticity of the harmful and hazardous drinker.
 On Articles 34 and 36, the Reference explained that, notwithstanding what had been said at the time of the Bill’s enactment, the Government were only relying on the aim of targeting harmful and hazardous drinkers as justification of the measure by reference to the application of Article 36. That was because taxation could achieve the wider aim of producing population-wide benefits. The Reference expressed a provisional view (para ) that the CJEU had “consistently found that the availability of excise duties… which are less disruptive of intra-EU trade and competition, mean that national rules fixing minimum… prices fail to meet the requirements for justification under Article 36…”. The Reference posed the following:
In the context of a justification sought under article 36 … where-
(i) a ... state has concluded that it is expedient in the interest of the protection of human health to increase the cost of consumption of a commodity – in casu alcoholic drinks – to consumers, or a section of those consumers; and
(ii) that commodity is one in respect of which the ... state is free to levy excise duties or other taxes (including taxes or duties based upon alcoholic content or volume or value or a mixture of such fiscal measures),
is it permissible under EU law, and if so under what conditions, for a ... state to reject such fiscal methods of increasing the price to the consumer in favour of legislative measures fixing minimum retail prices which distort intra EU trade and competition?”
 There was concern about the use of updated material, which had been available at the date of the hearing of the reclaiming motion but had not been before Parliament when the Bill was enacted. The petitioners maintained that compatibility required to be assessed at the time of the (appellate) court’s determination. Therefore the court should judge the derogation under Article 36 “on the basis of the best, current materials available to [the] national court at the time of the hearing of the case”. The Reference therefore asked:
Where a court … is called upon to decide whether a legislative measure which constitutes a quantitative restriction on trade incompatible with article 34 … may yet be justified under article 36 … on the grounds of the protection of human health, is that national court confined to examining only the information, evidence or other materials available to and considered by the legislator at the time at which the legislation was promulgated? And if not, what other restrictions might apply to the national court’s ability to consider all materials or evidence available and offered by the parties at the time of the decision of the national court?”
 The court was interested in the extent, depth or intensity of the judicial review. There was a dispute about the extent to which the court could proceed upon its own evaluation of the issues on the basis of the “evidence” before it and the degree to which it should allow the legislator a discretion in assessing whether an alternative measure was available and proportionate. Attention was drawn to the differing views (at that time) in Scotland and England (Sinclair Collis v Lord Advocate (supra) and R (Sinclair Collis) v Secretary of State for Health  QB 394). The following question was asked:
Where a court … is required, in its interpretation and application of EU law, to examine a contention … that a measure otherwise constituting a quantitative restriction within the scope of article 34 …is justified as a derogation, in the interests of the protection of human health, under article 36 …, to what extent is the national court required, or entitled, to form – on the basis of the materials before it – an objective view of the effectiveness of the measure in achieving the aim which is claimed; the availability of at least equivalent alternative measures less disruptive of intra EU competition; and the general proportionality of the measure?"
 The court was anxious about the effect of the restriction of the qualifying aim of the legislation, as stated by the first respondent, given that its wider benefits could not be prayed in aid where taxation could achieve them. The court accepted that taxation could result in a minimum price of 50p per unit, but noted that there could not be an exact equivalence between the benefits from minimum pricing and those from increased tax. The following was therefore asked:
In considering (in the context of a dispute as to whether a measure is justified on grounds of the protection of human health under article 36 …) the existence of an alternative measure, not disruptive, or at least less disruptive, of intra EU trade and competition, is it a legitimate ground for discarding that alternative measure that the effects of that alternative measure may not be precisely equivalent to the measure impugned under article 34 … but may bring further, additional benefits and respond to a wider, general aim?"
 The final question was directed towards the extent to which the court was entitled or required to attempt an assessment of the nature and degree to which the restriction affected intra EU trade and competition. The court questioned the first respondent’s view that, once it was accepted that a distortive effect would follow, its nature or extent did not matter. Accordingly, the CJEU was asked:
In assessing whether a national measure conceded, or found, to be a quantitative restriction in the sense of article 34 … for which justification is sought under article 36 … and in particular in assessing the proportionality of the measure, to what extent may a court charged with that function take into account its assessment of the nature and extent to which the measure offends as a quantitative restriction offensive to article 34?"
 On the common market issue, the Court expressed a provisional view (para ) that minimum pricing was “prima facie incompatible with the common organisation of the market where the market is organised on the basis of free formation of prices by normal market forces”, but the matter was a not “beyond debate”. It therefore asked the CJEU the following about what was by then an up-dated Regulation:
On a proper interpretation of EU law respecting the common organisation of the market in wine, in particular Regulation EU No 1308/2013, is it lawful for a ... state to promulgate a national measure which prescribes a minimum retail selling price for wine related to the quantity of alcohol in the sale product and which thus departs from the basis of free formation of price by market forces which otherwise underlies the market in wine?”
The CJEU Judgment ( 1 WLR 2283; 2 CMLR 27)
Opinion of the Advocate General
 The Advocate General (Yves Bot) produced his Opinion on 3 September 2015. Because much of its content (but not all) found its way into the judgment of the Court, it is not repeated at length. Some passages are worth noting. In summary, the Advocate General was of the view that minimum pricing did constitute an obstacle within the meaning of Article 34, since it deprived certain producers or importers of the commercial advantage that may result from lower cost prices. However, it could satisfy the principle of proportionality required under Article 36 if the national court held that (para 5):
“… it may reasonably be concluded on the evidence which the … State is required to place before the national court that the means chosen are appropriate for attaining the objective pursued and that, in making that choice, the… State did not exceed its discretion”.
The state required to take into account the extent to which minimum pricing impeded the free movement of goods, when compared with alternative measures that would enable the same objective to be attained. The court had to look at the material available when it determined the matter. The state could (para 7):
“… in order to pursue the objective of combating alcohol abuse, which forms part of the objective of the protection of public health, choose rules that impose a minimum retail price… that restricts trade… and distorts competition, rather than increased taxation of those products, only on condition that it shows that the measure chosen has additional advantages or fewer disadvantages than the alternative measure…”.
 A barrier to free movement could be justified, on one of the Article 36 public interest grounds, if it satisfied the conditions of proportionality. This meant that the measure had to be necessary to obtain the objective and not go beyond what was necessary to achieve it (para 72). The exercise of determining proportionality involved three stages: suitability for attaining the objective; necessity, that is the “minimum interference test”; and a balancing of the interests. In the context of a division of powers between the CJEU and the national court, the final assessment of proportionality was a matter for the referring court (para 81):
“which alone has jurisdiction to assess the facts of the case before it and to interpret the applicable national legislation. It is therefore ultimately for the national court to determine whether the national measure … is an appropriate means of ensuring the attainment of the objective … and whether it does not go beyond what is necessary to attain that objective. However, it is for the [CJEU] to provide the referring court with information to guide it in its interpretation, in particular as regards the criteria which it must take into account when forming its assessment.”
 Judicial Review of proportionality should be “marked by a certain degree of restraint” (para 82) because, first, it was for the state to decide on the degree of protection and the way in which it was to be achieved. States were allowed discretion in this area, representing the court’s concern not to substitute its own assessment for that of the state. Secondly, it was necessary to take into account the complexity of the assessment and any degree of uncertainty (para 84). Thirdly, it appeared to the Advocate General that (para 85):
“the somewhat experimental or fixed-term nature of [the legislation] is a factor to be taken into consideration by the … court, since it seems to establish in advance that [it] will be revised in the event that the reasons that led to [its] adoption have changed”.
It was nevertheless for the state to demonstrate proportionality. The reasons for the justification had to be accompanied by an analysis of suitability and proportionality. The degree to which the measure affected the free movement of goods must be taken into account when assessing proportionality (para 91). An assessment of whether there was a less restrictive measure had to be carried out. The national court had to identify the objective (para 119); in this case whether it was a two-fold, general and targeted, objective or only a targeted one.
 On suitability, the Advocate General expressed the view (para 127) that, when analysing any area of controversy, “it does not seem unreasonable” that minimum pricing should be considered appropriate to reduce consumption. According to the data, although minimum pricing might reduce consumption by hazardous and harmful drinkers, the effect would be much greater in the lower, rather than higher, income drinkers. The Advocate General was convinced (para 135) that the measure:
“meets the objective of combating alcohol abuse in a consistent and systematic manner by maintaining, in particular, that the measure forms part of a more general strategy of combating the harm caused by alcohol, including other measures such as the prohibition of specific promotional offers, and that the targeting of cheap alcoholic beverages may be justified by the fact that hazardous and harmful drinkers, including, in particular, the young, whose protection as a matter of priority is a legitimate concern, to a large extent consume that category of drinks.”
 In relation to necessity, and whether increased tax was a less restrictive measure, the Advocate General noted (para 145) the tobacco cases, in which the CJEU had held that fiscal legislation was an important and effective instrument for discouraging consumption. The objective of ensuring a high price level could adequately be attained by increased taxation, “the excise duty increases sooner or later being reflected in an increase in the retail selling price, without undermining the freedom to determine prices”. The Government required to show that increased taxation was not capable of meeting the targeted objective. They had not adduced “serious evidence” (para 149) to show that increased taxation would have a disproportionate impact. The Advocate General was:
“unable to see how that collateral effect of a general increase in taxes might be seen as negative… when … the scientific studies seem to show that hazardous or harmful consumption rises in line with the increase in consumer income, while … the quantity of the cheapest alcoholic beverages … falls with the increase in hazardous or harmful drinkers.”
Although it was for the national court to determine, it was difficult to justify minimum pricing which seemed less consistent and effective than increased taxation (para 151).
 On the single CMO Regulation, the Advocate General maintained (para 46) that it did not preclude national rules that present a minimum price, provided that those rules were justified by the objectives of the protection of health and of combating alcohol abuse, and did not go beyond that which was necessary in order to achieve those objectives.
 The CMOs had undergone a profound change, from their initial rationale of guaranteeing the income of farmers, to a liberal regime of prices determined in accordance with supply and demand. The establishment of the CMO did not prevent states from applying national rules which pursued an objective of general interest, other than those covered by the CMO, even if those rules were likely to effect the functioning of the common market in the sector concerned. The pursuit of a legitimate objective, such as the protection of health, was capable of justifying the action, even where a CMO had been established. The Regulation was not specifically intended to attain the objective of the protection of health in general or that of combating the dangerous or excessive consumption of alcohol. However, the principle of proportionality required the measure to meet the objective and not go beyond what was necessary to achieve it. The examination of proportionality had to be undertaken in the context of the Article 36 analysis.
Judgment of the Court
 On Articles 34 and 36, the CJEU, as the Commission had done earlier (supra), commenced its analysis by referring to the Dassonville formula (supra). The legislation prevented the lower cost price of imported products being reflected in the selling price. It was thus capable of hindering access to the UK market and was thereby a measure having the effect equivalent to a quantitative restriction (C-110/05 Commission v Italy (supra) at para 37). It could be justified on the grounds of the protection of human health and life if, but only if, the measure was appropriate for securing the achievement of the objective pursued and did not go beyond what was necessary in order to attain it. The CJEU continued (para 34):
“… it is apparent from the Explanatory Notes … and [the BRIA] that the legislation pursues a twofold objective, that of reducing, in a targeted way, both the consumption of alcohol by consumers whose consumption is hazardous or harmful, and also, generally, the population’s consumption of alcohol.”
The first respondent had confirmed that this objective, being aimed at the whole population, included moderate drinkers, albeit only secondarily.
 The CJEU was satisfied that the legislation did pursue, more generally, the objective of protecting health and life, which ranked foremost among the protections afforded by article 36. It was for states to decide what degree of protection they wished to assure (C-170/04 Rosengren v Riksåklagaren  ECR I-4071 at para 39). The CJEU added (para 36):
“As regards the appropriateness of that legislation to attaining that objective of the protection of human life and health, it must be observed that, … it does not seem unreasonable to consider that a measure that sets a minimum selling price of alcoholic drinks, the very specific aim of which is to increase the price of cheap alcoholic drinks, is capable of reducing the consumption of alcohol, in general, and the hazardous or harmful consumption of alcohol, in particular, given that drinkers whose consumption can be so described purchase, to a great extent, cheap alcoholic drinks.”
 A measure could be appropriate only if it genuinely reflected a concern to secure the attainment of the objective in a consistent and systematic manner (para 37). The measure was part of a more general political strategy which was designed to combat “the devastating effects of alcohol” (para 38). It thus constituted one of several measures “whose objective is to reduce, in a consistent and systematic manner, the consumption of alcohol by the Scottish population as a whole”. It followed that it appeared appropriate as a means of attaining the objective which it pursued (para 39).
 A separate analysis of proportionality did not have to be carried out for the Article 36 and Common Agricultural Policy (CAP) justification (para 40).
 The legislation could not benefit from an Article 36 derogation if human life and health could as effectively be protected by measures less restrictive of trade (Rosengren (supra) at para 43). There was a dispute as to whether taxation achieved effective protection. The objective of ensuring that the price of alcohol was high could be pursued adequately by increased taxation:
“… since increases in excise duties must sooner or later be reflected in increased retail selling prices … (see, by analogy, … Commission v Greece [(supra)]…para 31, and …[C-197/08] Commission v France [ ECR I-1599] … para 52).”
Taxation was liable to be less restrictive of trade than minimum pricing, because the latter significantly restricted freedom to determine retail prices. The fact that taxation entailed a general increase in prices, affecting moderate as well as harmful and hazardous drinkers, did “not appear, in the light of the twofold objective… to lead to the conclusion that such increased taxation is less effective…”. Indeed it may be more effective. The CJEU continued (para 49):
“It is however for the referring court, which alone has available to it all the matters of fact and law pertaining to the circumstances of the main proceedings, to determine whether a measure other than that provided for by the national legislation at issue in the main proceedings, such as increased taxation on alcoholic drinks, is capable of protecting human life and health as effectively as that legislation, while being less restrictive of trade in those products within the European Union”.
The answer to Questions 2 and 5 (para 50) was therefore that:
“… Articles 34 … and 36 … must be interpreted as precluding a … State choosing, in order to pursue the objective of the protection of human life and health by means of increasing the price of the consumption of alcohol, the option of legislation …, which imposes [a] MPU for the retail selling of alcoholic drinks, and rejecting a measure, such as increased excise duties, that may be less restrictive of trade and competition within the European Union. It is for the referring court to determine whether that is indeed the case, having regard to a detailed analysis of all the relevant factors in the case before it. The fact that the latter measure may bring additional benefits and be a broader response to the objective of combating alcohol misuse cannot, in itself, justify the rejection of that measure.”
 On Question 3, the CJEU stated that EU law had to be complied with at all relevant times, whether that was when the measure was adopted, when it was implemented and when it was applied to a particular case. The court was bound to assess the compatibility on the date when it gave its ruling (para 63), taking into consideration any relevant information, evidence or other material of which it had knowledge under the conditions laid down by the national law. Such an assessment was especially necessary where there was scientific uncertainty (para 64).
Question 3 was answered as follows:
“Article 36 … must be interpreted as meaning that the review of proportionality of a national measure, such as that at issue in the main proceedings, is not to be confined to examining only information, evidence or other material available to the national legislature when it adopted that measure. In circumstances such as those of the main proceedings, the compatibility of that measure with EU law must be reviewed on the basis of the information, evidence or other material available to the national court on the date on which it gives its ruling, under the conditions laid down by its national law.”
 The CJEU considered Questions 4 and 6 together. It re-iterated (para 52) that it was for states to decide on the level of health protection but, in doing so, they had to take into consideration the requirement for the free movement of goods in the EU. They needed to demonstrate proportionality; that the measure was necessary to achieve the declared objective and that that objective could not be achieved by “prohibitions or restrictions that are less extensive, or that are less disruptive of trade” (para 53, citing Rosengren (supra) at para 50). The reasons invoked “must be accompanied by appropriate evidence or by an analysis of the appropriateness and proportionality of the restrictive measure … and specific evidence substantiating its arguments” (para 54).
 However, the CJEU added that the “burden of proof cannot extend to creating the requirement that … [the state] must prove, positively, that no other conceivable measure could enable the legitimate objective pursued to be attained” (para 55, citing Commission v Italy (supra) at para 66). It was for the national court to review the legality; to assess the relevance of the evidence in order to determine the issue of proportionality. It must examine objectively whether the measure was “appropriate for the attainment of the objectives… and whether it is possible to attain those objectives by” less restrictive measures (para 56). Scientific uncertainty of a measure could be taken into account, as could the six year expiry limit in the legislation (para 57). The national court had to assess the nature and the scale of the restriction on the free movement of goods by comparison with other measures. It had to assess the effect on the CMO (para 58).
 Questions 2 and 4 were answered as follows:
“… Article 36 … must be interpreted as meaning that, where a national court examines national legislation in the light of the justification relating to the protection of the health and life of humans, under that article, it is bound to examine objectively whether it may reasonably be concluded from the evidence submitted by the … State concerned that the means chosen are appropriate for the attainment of the objectives pursued and whether it is possible to attain those objectives by measures that are less restrictive of the free movement of goods and of the CMO.”
 Returning to the First Question on the Single CMO (Common Organisation of Agricultural Markets) Regulation, the CJEU observed (para 17), contrary to the contention of the petitioners, that the Regulation “contains neither provisions that permit the fixing of the retail selling prices of wines … nor provisions that prohibit … States adopting national measures”. The “free formation of selling prices on the basis of fair competition is a component of the [Regulation] and constitutes the expression of the principle of free movement of goods in conditions of effective competition” (para 20). Minimum pricing could undermine competition, which was one of the objectives of the CAP. Consequently (para 24) minimum pricing was liable to undermine the Regulation as incompatible with the foundation principle of “free formation of selling prices of agricultural products on the basis of fair competition”. However, the establishment of a CMO did not prevent the application of rules intended to attain an objective relating to the general interest, even if they affected the functioning of the market (para 26). The objective of the protection of human life and health could be used to justify a measure which otherwise undermined the system (para 27), provided it satisfied the test of proportionality; that is that it “must be appropriate for attaining the objective pursued, and must not go beyond what is necessary to attain that objective” (para 28).
 The Court answered question 1 as follows:
“… the Single CMO Regulation must be interpreted as not precluding a national measure, such as that at issue in the main proceedings, which imposes [a] MPU for the retail selling of wines, provided that that measure is in fact an appropriate means of securing the objective of the protection of human life and health and that, taking into consideration the objectives of the CAP and the proper functioning of the CMO, it does not go beyond what is necessary to attain that objective of the protection of human life and health.”
The New Material
 The written pleadings make no reference to material available later than 2012. At first instance, there had been an interlocutor of 9 October 2012 appointing parties to lodge all productions by 16 October. Some productions were received late, but otherwise the petition and answers proceeded to a hearing on the basis of the pleadings and documents then lodged. That hearing took place in January 2013.
 The Lord Ordinary had been able to consider the Policy Memorandum, the Explanatory Notes, the BRIA, the Parliamentary Debates, the Opinion of the Commission and those from the various EU states. He had the University of Sheffield “Model based appraisals” from 2009 to 2012, certain critiques and appraisals thereon, notably papers from Duffy and Snowden (eg The Minimal Evidence for Minimum Pricing: the fatal flaws in the Sheffield Alcohol Policy Model (2012)) and the Yarrow and Decker report (supra) for the petitioners. A copy of the Wagenaar report (supra) was available as were extracts from the Scottish Health Survey 2011.
 The new material sought to be introduced at the pre-Reference stage of the reclaiming motion included three NHS MESAS documents, namely the Second Annual Report (2012), the impact of the Alcohol (Scotland) Act on off-trade alcohol sales in Scotland (2013), and the annual update of alcohol sales and price band analyses (2013). A further NHS document, namely The price distribution of wine from different countries of origin sold in Scotland’s off-trade (2013) was available. A new Sheffield modelling paper was produced (Modelled income group – specific impacts of alcohol minimum unit pricing in England 2014/15: Policy Appraisal using new developments etc (2013)). Two papers from JP Nelson (Does Heavy Drinking by Adults Respond to Higher Alcohol Prices and Taxes? A Survey and Assessment (2013) and Meta-analysis of alcohol price and income elasticities – with corrections for publication bias (2013)) were lodged along with a few other articles. The WHO’s Status Report on Alcohol and Health in 25 European Countries (2013) was produced as was a Eurostat news release (Consumer Price Levels (2013)) and HM Revenue & Custom’s Working Paper 10 “Econometric Analysis of Alcohol Consumption in the UK”. The narrative in the Reference (supra), incorporates some of this additional material. It had been lodged in accordance with an interlocutor of 29 November 2013, which permitted “supplementary materials” to become part of the Appendix to the reclaiming print.
 By the time of the final hearing, there had been no decision about what materials could be looked at beyond those before the Lord Ordinary. In this context, it is important to observe that, very often in judicial review proceedings, the court is not looking at evidence (ie testimony or agreed fact) but information, contained in documents which state facts which may be disputed. The provenance of the material may be of variable quality. Prior to the Reference, it was the petitioners who seemed keen for the court to look at material which had not been before the Lord Ordinary. In making the Reference, the court did this, but there was no view expressed on what might be considered in the final analysis. The Reference was simply that. It did not involve the court making any decisions, other than one’s stated to be provisional, on the reclaiming motion itself. After the Judgment of the CJEU had been received, it was the first respondent who sought to introduce new material. This time the petitioners opposed the application on the basis that the matter had been decided in terms of the interlocutor of 29 November 2013, which had imposed a deadline of 6 December for the lodging of materials.
 The new material consists of, first, up to date NHS MESAS documents, notably the Annual update of alcohol sales and price band analyses (2015) and the Final Report (2016). Another Sheffield University paper (Model-based appraisal of the comparative impact of Minimum Unit Pricing and taxation policies in Scotland (2016) was made available, as was a paper from the Sheffield Alcohol Group (Meier et al; Estimated Effects of Different Alcohol Taxation and Price Policies on Health Inequalities etc (2016). There were new articles from, amongst others, Prof Ludbrook (supra). Some Scottish Government and HM Revenue & Customs publications were produced along with material emanating from the first petitioners. There was a chapter from the House of Lords European Union Committee on Taxation and the Irish equivalent of the BRIA.
 The petitioners responded by tendering extracts from the Scottish Health Survey 2014, further publications emanating from the first petitioners, sundry statistics from the UK Government and abroad on taxation and other matters and, inter alia, the Scottish National Party manifesto. Not all of this new material was ultimately founded upon. In so far as it was, it is included in the narrative of the submissions (infra).
 The petitioners argued that the “cut off” date, for the purposes of the reclaiming motion, remained 6 December 2013. Any material after that fell to be excluded. The same rules applied as were appropriate when the court, in rare circumstances, ordered additional proof during an appeal process on the grounds of res noviter veniens ad notitiam (information newly discovered; Rankin v Jack 2010 SC 642, Lord Reed at para ). Finality in litigation was important (ibid) and an appeal on the basis of res noviter could only proceed where the evidence was not, and could not have been, known to the party at the time of the proof (ibid at para ).
 The first respondent submitted that the view of the CJEU was that the court should approach the matter on a contemporary basis. Although the principle of finality was important, the court retained a discretion to allow additional proof (ibid paras [36-37]) where the interests of justice required the principle of finality to be overridden (ibid para ). The proceedings had been running for more than four years. The Lord Ordinary had considered material which had not been before the Parliament, although he was of the view that it did not alter matters. These were extraordinary proceedings where the legislation was not yet in force. Parliament would need to be presented with new material before making the Order. Both the nature of the proceedings and the cogency of the new material justified its admission. It would be contrary to the interests of justice for the court to assess the necessity of legislative intervention by reference to outdated statistics in conflict with current trends.
 There were, according to the first respondent, several reasons for considering the new material. The first was in case the court misunderstood the position. In 2014, at the time of the Reference, the data available had been from 2012, when the trend in consumption was down. There had been a decline in mortality and morbidity. The petitioners had therefore argued that the measure was unnecessary. The trend had now flattened as was illustrated in the new MESAS survey (Annual update and alcohol sales and price band analyses (August 2015)). This stated that, in 2014, 10.7 litres of pure alcohol were sold per adult (20.5 units per adult per week). There had been a “levelling off in the recent downward trend”. The MESAS Final Annual Report (March 2016) repeated this. It noted (para 2.3) that “alcohol-related mortality and hospitalisations remain high and mortality rates have not declined since 2012 … although it is too early to say whether this marks the start of a long-term change in trend”. This report had concluded (para 7) that:
“Despite these recent improvements, Scotland continues to experience substantially higher levels of alcohol-related mortality and morbidity compared to the 1980s and compared to England & Wales. Inequalities in alcohol-related harm persist. Those living in the most deprived areas, especially men, continue to experience the highest levels of alcohol-related morbidity and mortality. There is, therefore, a continued need for action to further reduce alcohol-related harm in Scotland and to address these health inequalities. MUP has not been implemented and this is likely to have constrained the strategy’s contribution to declining alcohol consumption and .related harm, and efforts to implement should continue. There was no decrease in either alcohol consumption (sales) or alcohol-related mortality in 2013 and 2014. While it is too early to say if this marks a change in trends, continued monitoring is required.”
The information produced by the first respondent had been based on sales. The petitioners had relied on self-reported consumption, which was recognised to be a less accurate view.
 The second reason was that the new material included the answers of the Government to questions posed by the CJEU prior to its judgment. The answers were no more than an up-dated repetition of the contents of the BRIA. Thirdly, there was a further report from Sheffield University (Model-based appraisal of the comparative impact of Minimum Unit Pricing and taxation policies in Scotland (April 2016)), which addressed certain issues identified as relevant by the CJEU.
 The normal rules in relation to the consideration of material in the course of first instance and appellate proceedings are clear. They differ at the first instance stage depending upon the type of process involved. Thus in an Ordinary action, where the case is to be determined after a formal proof, the relevant productions will have been lodged prior to the proof according to the applicable rule (RCS 36.3) or by leave of the court. A reclaiming motion against a Lord Ordinary’s determination will not normally involve a consideration of factual material not properly before the Lord Ordinary at the proof. The exception is where the principle of res noviter veniens ad notitiam applies to permit new evidence capable of proving a material new fact. That was the context of Rankin v Jack 2010 SC 642. The principle of finality will often prevail, where the effect of the application is to re-open the proof.
 Petitions generally fall into a different category, in that they are summary applications to the court, in which the procedure is highly flexible. Petitions for judicial review are a hybrid. They are not governed by the same general procedural rules as those in Ordinary procedure. “Speedy” determination is the key (RCS 58.11.(2)) and any “evidence” may take the form of an affidavit (RSC 58.11.(2)(j)). One obvious point, for present purposes, is that such petitions do not culminate in a proof, requiring evidence to establish fact. Rather, a “substantive hearing” will take place, although in certain situations testimony may be heard at that hearing. Any documents to be founded upon ought to be lodged, or at least identified, at the outset (RCS 58.3.(4)). The Lord Ordinary may set out the facts upon which he has proceeded, but these will not normally be based on testimony or other formal proof. In that context therefore, the concept of finality, as it applies to a decision following upon a proof, is different.
 The rules in reclaiming motions are broadly the same in their application to interlocutors in Ordinary actions as they are to those after Judicial Review petition hearings. In each, the reclaimer (and possibly also the respondent) may lodge an Appendix containing the documents which he wishes to be considered by the appellate Division (RCS 38.19.(8)). The reproduction of documents in an Appendix does not confer on them any special status as productions.
 The court can, and in this case did, vary the timetable for lodging an Appendix by providing a specific date prior to the appeal hearing. It stated that each party was to lodge, by 6 December 2013, “any supplementary materials upon which they wish to rely” for inclusion in the Appendix. The parties took advantage of this and lodged documents which they wished, and presumably still wish, considered. They were referred to at the initial hearing of the reclaiming motion in February 2014. However, that motion did not result in a determination. It resulted in the Reference. As a consequence, the motion has proceeded to a continued hearing, with the benefit of the CJEU’s answers, over 2 years later. In advance of that, by interlocutor dated 28 January 2016, the court again allowed parties to lodge new material upon which they wished to rely. That was without prejudice to the issue of whether, ultimately, the court should take that material into account.
 The issue is not just whether the court should look at post Reference material. It is whether, in determining the reclaiming motion, it should look at any material post dating the Lord Ordinary’s Opinion in May 2013. Although a timetable was set for the lodging of new material in November 2013, the setting of a deadline does not preclude the court from allowing further material to be incorporated in an appendix, if a reason for late inclusion is acceptable. A change of circumstances, including the fact that the material is newly available, is one obvious example. Be that as it may, inclusion in an Appendix or permission to lodge it otherwise, does not determine the question of whether the court should take it into account.
 The CJEU’s advice, that compatibility must be reviewed on the basis of the information available to the national court on the date when it gives its ruling, is primarily applicable to the point at first instance when the Lord Ordinary hears the case prior to issuing his decision. What happens in an appellate process will depend upon the domestic rules applicable upon appeals (ie “the conditions laid down by … national law”; CJEU Answer to Question 3). The norm in reclaiming motions is that they are reviews of the decision of the Lord Ordinary based upon the material presented to him at the time of the substantive hearing. However, the court is entitled to have regard to new material where it considers, in its discretion, that the interests of justice require that it be taken into account. At the pre-Reference stage, the petitioners were content to attempt to introduce new material without advancing a res noviter case. The first respondent was at that time maintaining, contrary to the CJEU judgment, that only information before (or at least available to) the Parliament at the time of enactment ought to be considered. Prior to the final hearing, the first respondent sought to introduce new material and this time the petitioners objected on the basis of an absence of res noviter.
 Having regard to the time which has expired between the Lord Ordinary’s interlocutor and the final hearing of the reclaiming motion, it is in the interests of justice that any new material, which is pertinent to the issues of Scotland’s alcohol consumption and the effect which minimum pricing may have on it and intra -EU trade, should be considered. Quite apart from anything else, both the Reference and the CJEU’s judgment involved a consideration of new material and it would seem strange if that material were not to be taken into account now. Once it is accepted that the Lord Ordinary’s Opinion ought to be considered in light of that material, which was not before him, it must be in the interests of justice to look at any additional material which effectively amounts to an update or revision of that material or was information requested, or taken into account, by the CJEU.
 In so deciding, it is of some note that the first petitioners are a trade association to whom the material will come as little surprise and who have the resources to deal with it in such a manner as they might deem fit. There is no unfairness in the granting of the application to consider all of the new material; the petitioners having had sufficient time to consider, and react to, it. It would not be in the interests of justice for the court to deal with the issues by proceeding upon data which is substantially out of date.
 That is not to say that, in considering the new material, the court is thereby opening up the Lord Ordinary’s decision for reconsideration as if this court was one at first instance. If the Lord Ordinary’s assessment of the import of the information before him cannot be faulted in any substantial way, the new information can only be regarded as significant if this court were to hold that it would have altered his view of the facts. This court must keep its own function as an appellate body firmly in mind and refrain from reassessing the issues, as appeared to be the desire of the petitioners, from scratch and without due deference to the Lord Ordinary’s findings.
The Appellate Context and Submissions
Grounds of Appeal
 The parties agreed that the respondents should address the court first. The court did not demur to that course of action. The basis for the first respondent acceding to this was an acceptance that it was for the Government to satisfy the court that the legislation was EU compliant notwithstanding the breach of Article 34. It nevertheless remains the position that this is a reclaiming motion from the Lord Ordinary. Although the Government do have an obligation to demonstrate EU compliance, that is a matter which looms large at first instance. The petitioners’ grounds of appeal, which remain unaltered, are that the Lord Ordinary erred in certain specific respects. Returning to a theme developed earlier, it is important that the appellate court does not attempt to rehear the entire case, as seemed to be implied by some of the submissions, as if it were a court of first instance. The appeal is circumscribed by procedural rules and practices confining the scope of review of the Lord Ordinary’s decision within reasonable bounds.
 The grounds of appeal are, in summary, first (ground 1) that the Lord Ordinary erred in identifying the correct tests in considering proportionality and applying these tests to the legislation. He had failed (ground 1(a)) to identify the correct aim of the legislation. This was not, as the Lord Ordinary had thought, the reduction of alcohol consumption generally and that of harmful and hazardous drinkers in particular. There was no evidence to show that the reduction in general consumption resulted in a significant degree of protection to health and life. The legislation would impact on persons who drank moderately and thus did not target harmful and hazardous drinkers. If the objective was to make low-priced drinks less competitive, that was not legitimate. The Lord Ordinary had also erred in taking into account what he said were other consequential (largely public order and economic) benefits accruing to Scotland. There was no, or insufficient, evidence of this.
 The Lord Ordinary had erred (ground 1(b)) in his determination that minimum pricing was necessary to secure a legitimate aim. He had failed to “exercise the due degree of critical scrutiny” of the evidence. That evidence showed that harmful and hazardous drinking was associated more with affluence than poverty. The Lord Ordinary failed to give due weight to the submission that minimum pricing would therefore “be unlikely to affect the … more affluent harmful and hazardous drinkers” who were among those supposedly targeted. The first respondent had changed the target, from that stated at the stage of the Bill, to poor harmful and hazardous drinkers. This demonstrated a failure to approach the issue in a consistent and systematic manner. It was not legitimate to press upon the court a justification which differed from that before Parliament.
 The Lord Ordinary had erred (ground 1(c)) in his consideration of whether a measure less distortive could have been used. He had failed to have proper regard to the extent to which legislation was needed at all, given that other measures were already achieving “benefits in relation to drinking”. The Government had presented no evidence about the effects in Scotland or the EU and had accepted that tax increase would be less distortive. The Lord Ordinary had erred in holding that, nevertheless, the Government were still entitled to adopt measures which, in their view, achieved the desired aim without taking into account these effects. His “dismissal of the excise duty alternative was peremptory and ill-founded” (para 1.8.3). He had erred in holding that any alternative measure had to have “precisely the same effect” (para 1.8.4). Adjustment of taxation and a ban on “below cost” sales was clearly an alternative way of ensuring the same retail price.
 The Lord Ordinary had failed (ground 1(d)) to consider and determine whether the distortive effect of the measure in the EU was outweighed by the benefits in Scotland. He had erred in considering that this was within the margin of appreciation.
 The Lord Ordinary had failed (ground 2) to give the EU Commission’s Opinion sufficient respect and weight. He had acted incompatibly “with his duties of sincere co-operation with the Commission” (para 2.2.3, citing Article 4(3) of the Treaty of the European Union). Instead of dismissing the Commission’s Opinion, he ought to have made a Reference to the CJEU. His refusal to do so was “contrary to his duties” (para 2.3).
 The Lord Ordinary had failed (ground 3) to have proper regard to the case law of the CJEU, notably recent tobacco cases. He had failed in rejecting a contention that the intensity of the application of the proportionality test depended upon context and ought, in this case, to be “high” (para 3.2). His according to the Government a generous degree of discretion was misplaced. None of the material relied upon suggested that the first respondent or the Parliament had given substantial consideration to the excise duty alternative.
 The Lord Ordinary had erred (ground 4) in rejecting the petitioners’ arguments about the CAP. The single CMO formed a complete system, especially as regards price regulation. States could not unilaterally adopt measures, such as minimum pricing, which ran contrary to the CAP aims, notably to ensure that supplies reached consumers at reasonable prices. The general principles of the CAP precluded measures purporting to fix or determine the prices for wine. In upholding the measure, the Lord Ordinary had failed to enforce the obligation on states to refrain from taking any measure which would undermine, or create exceptions to, the CMO.
 The final ground (ground 5) was that the Lord Ordinary had erred in holding that the legislation was incompatible with World Trade Organisation requirements. The legislation was a violation of Article III.4 of the General Agreement on Tariffs and Trade 1994, and hence of EU law, in that imports from outside the EU would receive less favourable treatment than competing domestic products. This ground was not insisted upon.
 The first respondent submitted that the CJEU judgment set out the structure within which proportionality, in the context of Article 36, required to be considered. Minimum pricing would be proportionate to pursue the objective of the protection of life and health provided that the national court, having carried out an objective assessment, was satisfied of two conditions. First, it must reasonably be concluded, from the evidence submitted by the Government, that minimum pricing was appropriate to achieve the objective of life and health protection. Secondly, it must not be possible to attain that objective by an alternative means which was less restrictive of the free movement of goods and of the Common Organisation of agricultural Markets.
 The case for minimum pricing, as contained in the BRIA, should be considered in four stages, which corresponded to the requirements of the proportionality test: first, the objective of the measure; secondly, the scale of the alcohol problem; thirdly, the suitability of minimum pricing to tackle alcohol related harm; and fourthly, the necessity to take any action, and, if so, whether minimum pricing was the least restrictive way of doing so.
 The objective, as set out in the BRIA, was to protect health and life in accordance with Article 36. There were dual aims. The first, and principal, aim was to target hazardous and harmful drinkers, who were most at risk from the dangers of alcohol consumption, and who tended to drink alcohol that was cheap relative to its strength. The second objective was to have a positive effect on the health of the population as a whole.
 Scotland had a serious problem with alcohol. The average amount of pure alcohol sold per adult in Scotland in 2014 (10.7 litres, supra) was 18% more than that in England & Wales (MESAS, Annual update (2015), p 4). The overall mortality rate was 49% higher than in 1981 and in 2014 was 5% higher than in 2012 (MESAS, Final Report (2016) (supra), para 2.2.1; App B). There was a wealth of evidence demonstrating a strong relationship between affordability and alcohol-related harm. As affordability increased from the early 1980s, so had harm.
 The high consumption was driven by the availability of cheap alcohol. Some 52% of alcohol sold in the off-trade in Scotland in 2014 retailed at less than 50p per unit (60% in 2012; MESAS, Annual update (2013) (supra) p 19). The volume was substantially higher than in England and Wales. Most (62%) of the additional alcohol sold in Scotland (as compared to England & Wales) retailed at less than 50p per unit, with 71% being cheap spirits (MESAS, Annual update (2015), pp 21-22). Some 2.4 times more vodka was sold at under 50p per unit in Scotland when compared with England & Wales (ibid p 22).
 Hazardous and harmful drinkers were at the greatest risk of harm. Combined, they accounted for only 25% of the population, yet they consumed 71% of the alcohol (Sheffield University: Model-based appraisal etc (2016), para 4.1.1 p 38). Whilst there was a greater proportion of hazardous and harmful drinkers in higher income groups, the degree of harm suffered as a consequence depended upon the quantity of consumption, and not simply the crossing of the hazardous or harmful statistical threshold. Those in the lowest income group were at the greatest risk of harm, because on average they drank more. Average weekly consumption by harmful drinkers was considerably higher in lower income communities. Male harmful drinkers in the lowest income group drank an average of 112 units per week (MESAS, Annual Report (2013) Fig 4.23). This was more than 30 units higher than harmful male drinkers in other income groups. The same pattern existed for females. Hospital stays were on average 7.7 times higher for harmful drinkers in the most deprived quintile (MESA, Final Report (2016), App C), and the death rate was almost 8 times higher (MESA, Final Report (2016), section 2.1.1 Fig 9B, App B).
 Minimum pricing was a suitable measure to tackle alcohol related harm. The majority of the Canadian provinces have had minimum pricing for many years. The empirical evidence suggested a direct link between changes in minimum price and changes in consumption and harm in general (BRIA, para 2.35). The Sheffield University model had consistently predicted that minimum pricing would reduce harm generally, but target in particular those who drank at hazardous and harmful levels. The policy would reach full effect after 20 years, when there would be a reduction of 121 deaths and 2,042 hospital admissions per annum. By year 5 following implementation, 77% of the full impact on deaths, and 95% of the full impact on hospital admissions would be achieved (Sheffield University: Model-based appraisal etc (2016), para 4.2.4).
 The court should apply an objective approach when considering whether minimum pricing was a suitable measure. It should consider what conclusions could reasonably be reached on the evidence presented by the first respondent. This was confirmed by the Opinion of the Advocate General. It reflected the need for a “certain degree of restraint” in the exercise of judicial review. This was consistent with the state having a discretion on how to achieve the objective of protecting life and health. The approach, which the Lord Ordinary had taken, followed that in Sinclair Collis v Lord Advocate (supra). It was appropriate to take into account scientific uncertainty being mitigated by the sunset clause.
 Two issues on suitability had been raised by the petitioners; first, elasticity (infra), and secondly, the potential for a greater impact on less affluent drinkers. The CJEU had considered both points. It had still reached the conclusion that the legislation appeared to be a suitable means of attaining the objective which it pursued. It recognised that there were competing views on the suitability of the measure. Within the state’s discretion, it was not unreasonable to prefer one body of evidence to the other, so long as the information which supported the choice was cogent. The court should thus consider objectively whether the information relied upon by the Government was cogent. If it was, it was reasonable for the Government to consider the measure suitable. The court should not interfere with that view. It was not appropriate for the court to substitute its own view for that of the Government.
 Price elasticity was the measure by which drinkers would adjust their consumption in response to an increase in price. The Sheffield University model was conducted with the input of an analysis of over 100 studies (Wagenaar et al, 2009, (supra)). The WHO’s Global Alcohol Strategy to reduce the harmful use of alcohol (2010), which was agreed by all Member States, had accepted that consumers of alcohol, including heavy drinkers and young people, were sensitive to changes in price (para 32, p 16). The latest reports confirmed this. The Sheffield Report used the elasticity data contained in the Living Costs and Foods Survey (Sheffield University: Model-based appraisal etc (2016), para 3.2.6 and Table 3.2). The petitioners relied upon the Yarrow & Decker reports (supra), which used HMRC elasticity data. Applying the HMRC elasticities, the estimated impact would be greater (Sheffield University: Model-based appraisal etc (2016), para 4.4, Table 4.24). There was no merit in the petitioners’ criticism of the elasticities used in the Sheffield modelling.
 The petitioners relied upon a paper by Neilson (supra), which concluded that the link between alcohol consumption and price was uncertain, and that heavy drinkers were immune to an increase in price. That paper was a review of prior documentation, and was not in itself empirical research.
 The petitioners’ view, that minimum pricing would have the greatest impact on those living in poverty, whilst being ineffective in targeting more affluent drinkers, where there were a higher proportion of hazardous and harmful drinkers, was erroneous. First, the BRIA recognised that it would have a greater impact on those in poverty. Whilst there was a higher proportion of hazardous and harmful drinkers among the affluent, the proportion of hazardous and harmful drinkers in poverty consumed more alcohol. The pattern of consumption was different. Secondly, the risk of morbidity and mortality through alcohol consumption was considerably greater for those in poverty. In reducing consumption, with the consequent decrease in alcohol-related illness, minimum pricing was likely to have a greater impact on the health of those living in poverty.
 It was legitimate to attempt to attack the harm suffered by hazardous and harmful drinkers in poverty. Much of the problem with alcohol consumption in Scotland was caused by cheap alcohol, notably vodka. It was not possible to tax vodka differently from whisky or other spirits. The heavier the drinker, the more units were purchased at below 50p (Sheffield University: Model-based appraisal etc (2016), Fig 4.6 and para 4.1.1). The BRIA recognised that those in lower income quintiles would be disproportionately affected (BRIA, para 2.29). Tackling cheap alcohol was the aim. Minimum pricing would achieve a benefit across all income groups. Any criticism of the Sheffield Report, based on the fact that it compared those in poverty with those who were not, as compared to a breakdown of harm by quintile, was unfounded. The population of Scotland was too small for the figures to be meaningful, when broken down in that manner.
 The Lord Ordinary had dealt with the petitioners’ arguments correctly. Minimum pricing targeted the harm caused by alcohol. This was suffered disproportionately by those in the lowest income quintiles. The fact that the policy would therefore have a greater impact on those in the lowest quintiles was a positive. It produced a result that other policies would fail to achieve. The CJEU had been satisfied that minimum pricing was a suitable measure to pursue the objective. The first respondent had updated the material presented to the CJEU in the form of the Sheffield University Report. The position had been consistently stated in the BRIA, before the Lord Ordinary, before the CJEU, and before this court. The Act and the draft Order were appropriate and suitable as a means of achieving the objective.
 The final stage was the necessity of minimum pricing. There were two questions. First, was it necessary for the first respondent to take any action? Secondly, was minimum pricing the least restrictive option? Minimum pricing was part of the Government’s wider alcohol strategy. Before the Lord Ordinary, the first respondent had argued that minimum pricing had the potential to produce a step-change in the rate of improvement of health as a result of the decline in consumption. The updated statistical information confirmed that, since 2012, consumption had risen, with a consequent increase in mortality. The cumulative impact would be to save 58 lives in the first year, and 2,036 lives by year 20 (Sheffield University: Model-based appraisal etc (2016), at Table 4.16). The delay in implementation had constrained the impact of the wider alcohol strategy (MESAS, Final Report (2016), p 8).
 The crucial issue was whether it could reasonably be concluded from the evidence produced by the Government that it was not possible to attain the objective by measures which were less restrictive of the free movement of goods and the CMO. The first respondent did not need to establish that no other conceivable measure could enable the objective to be achieved.
 It was accepted that minimum pricing would have an impact on the market of alcohol generally, including wine. It would affect domestic products, as well as EU imports. It was not possible to predict the particular effect on different sectors of the market. The BRIA had noted that there had been no consistent response from retailers when they had been questioned on that point (BRIA, paras 114-118). The second Yarrow and Decker report, cited three expected economic effects: (i) discrimination in favour of low cost products; (ii) discrimination against low cost suppliers; and (iii) an incentive on the part of retailers to sell low cost products. The last two propositions were contradictory. The question was whether retailers would continue to stock low cost products because of the higher margin, or whether they would reduce the volume of lower priced stock and concentrate on better quality products. Parliament had not received a consistent response when they had asked this question. Put simply, there was no consensus in the industry. The court was entitled to take the existence of the sunset clause into account when considering scientific uncertainty, as well as the fact that minimum pricing was an unique experiment, which had not been trialed in this form in any other country.
 All that could be done was to consider the range of products that might be affected. Within that range, wine was the category with the lowest percentage of the market sold at less than 50p per unit. Most of the impact would fall on non-EU imported wine. Of the 15 wines with the largest value sales in the off-trade in 2013, none were produced in EU countries (Scottish Grocers Federation, 2014, Off-trade – Scotland’s most valuable brands, p 64). The case for minimum pricing was made for all types of alcohol, or none, because of the possibility of trading down. It did not make sense to treat different types of alcohol differently. From the CJEU’s answer to Question 1 of the reference, there was no need to treat wine differently from any other type of alcohol.
 The reasons for rejecting tax increases as an option were summarised in the BRIA (paras 45-47). The UK already had the fourth highest rate of duty on alcohol in the EU, yet the Scottish population continued to suffer the harm which it did. Taxation did not target alcohol which was cheap relative to its strength. Excise duty could not address product value and VAT could not address product strength. Directive 92/83/EEC prevented duty being set in proportion to the percentage of pure alcohol. This meant that wine required to be taxed at the same level, when the ABV could be anywhere between 8.5% and 15%.
 The effectiveness of minimum pricing could be compared with tax measures by either calculating the rates of excise duty that would have to be imposed to achieve a 50p per unit minimum price, or by calculating the level of increase in the rates of tax which would produce the same health benefits as minimum pricing. This was illustrated as follows:
Examples of duty rates required to ensure products retail at a MUP of 50p
UK duty rate 2012/13
Duty rate required to ensure all products retail at a minimum of 50p per unit
£ per hectolitre per cent of alcohol in the beer
£ per hectolitre of product
Cider and perry: 1.2% - 7.5% abv.
Wine 8.5% - 15% abv.
Sparkling wine: 8.5% - 15% abv.
£ per litres of pure alcohol
Impact of 50p MUP and possible tax increase on typical products
Retail price as at November 2012 (£)
Impact of a Tax- based approach (£) (% increase)
Impact of 50p Minimum unit price (£)
Low-price brand 4x440mls
Mid-price brand 4x440mls
High-price brand 500mls
Low-price brand 2000mls
Mid-price brand 4x440mls
High-price brand 500mls
| || || || || || || |
 Tax increases of between 27% and 70% would be required to achieve the same public health benefits as a 50p per unit minimum price (Sheffield University: Annual Report (2016), section 4.3). Tax increases could not remove the ability to trade down, so there was greater uncertainty about the impact of increased tax on hazardous and harmful drinkers. The petitioners’ reliance on the English Sheffield University modelling was misplaced, as the differences in the level and distribution of consumption, preference for cheap alcohol, and patterns of harm all distinguished Scotland from England. If the tax and VAT rates were set as high as they required to be to achieve the same effect as minimum pricing, that would itself be distortive of the market. The Yarrow & Decker reports assumed that tax increases would be less restrictive of the free movement of goods, without considering the level of increase required to achieve the same effect. The petitioners had submitted in the past that the structure of excise duty in the UK distorted competition. Whisky was taxed more heavily than other alcohol. If tax increases were used as an alternative, this would exacerbate the problem. Increases in tax could not target cheap vodka. The trade impact of increasing taxation by even 27% would be entirely disproportionate to the aim of the legislation.
 The level of tax increases required would be so disproportionate and inconsistent with the aim of the legislation that they were not a realistic alternative. Parliament could not sensibly propose increases at that level. Increasing the price of alcohol and disadvantaging all consumers, including those who were not at harm and would reap no commensurate health benefit, was not the objective of the legislation. If minimum pricing was unlawful, the Government would have no practical means to save the 2000 lives which the measure was predicted to save in the first 20 years alone. With no realistic and viable alternative to minimum pricing, the right to trade, under Article 34, was weighed directly against the duty of the Government to protect the lives and health of its citizens, under Article 36. The protection of life and health of citizens ought to be given precedence. It was not possible for the Government to attain its objectives by measures which were less restrictive of the free movement of goods.
 The second respondent confined his submissions to the approach that the court should take to reviewing the legislation, in light of the judgment of the CJEU. The court ought to assess the proportionality on the basis of the information which was available at the time of the (appeal) hearing.
 The Advocate General of the CJEU had noted that the national court should exercise a certain degree of restraint in carrying out a judicial review of the legislation. The CJEU recalled that the protection of the health and life of humans ranked foremost among the interests protected by Article 36. It was for the first respondent to adduce evidence that the public health justification was met. The burden of proof did not extend to proving positively that no other conceivable measure could enable the objective to be obtained under the same conditions (C-110/05 Commission v Italy (supra)). The court required to consider objectively whether it could reasonably be concluded from the evidence produced by the Government that minimum pricing was appropriate to attain the objective, and whether it was possible to attain the objective by a measure which was less restrictive. The court was entitled to take into account the extent of scientific uncertainty. The CJEU had had no difficulty in concluding that minimum pricing was suitable to achieve the objective. The necessity assessment had to be undertaken with regard to the objectives of the CAP and CMO. It was for the court to review the necessity aspect.
 The CJEU did not agree with its Advocate General’s view that proportionality had to be assessed at a third stage, which required the assessment of proportionality as a whole. There was no third stage balancing exercise in a public health case. Any such exercise, which was ultimately an attempt to weigh high-level policy objectives against each other, was difficult to reconcile with the margin of discretion. The court could not second-guess the balance which had been struck by the Government between policy objectives. If the court did require to conduct such a balancing exercise, it would be almost impossible to conclude that the protection of public health ought not to prevail.
 Minimum pricing breached EU law because it had the effect of limiting the freedom of manufacturers and importers to set their prices in normal competitive conditions. As a result, intra-EU trade was disrupted. The measures would cancel out any potential competitive advantage of low-cost producers. They would have an adverse impact on the proper operation of the EU’s Common Market Organisation in wine. The court had to find that minimum pricing was incompatible with EU law, and so “not law”, unless the first respondent could demonstrate that the interference was proportionate. The first respondent had failed to discharge the onus on him to show that minimum pricing was a proportionate interference with (i) the free movement of goods; and (ii) the objectives of the CAP and the proper functioning of the CMO. He had failed to establish that alternative pricing measures, including a combination of one or more of: excise duty; additional sales tax; a ban on below cost sales of alcohol; and a ban on sales with an artificially low profit margin, were less effective in achieving the objective. Those measures were less disruptive of free trade and less distortive of competition, as well as being specifically less disruptive and distortive of the market in wine, and therefore more compatible with the objectives of the CAP.
 The role of the court was not to carry out a review of the reasonableness of Parliament’s decision in passing the legislation, but rather to consider all of the evidence and material now before it and to decide whether the legislation complied with the principle of proportionality. The court had to consider three questions. The first was whether it could reasonably be concluded from the evidence that imposing a minimum price was an appropriate means of securing the objective. Secondly, taking into account the objectives of the CAP and the proper functioning of the CMO in wine, did minimum pricing go beyond what was necessary to achieve the objective? Finally, was it possible to attain the objective by measures which were less restrictive of the free movement of goods?
 Minimum pricing had to be necessary. The alternative measures, which formed part of the broad approach taken by the Government, were effectively tackling the issue. Health statistics indicated that, since 2003, there had been a steady and significant decrease in over-consumption and misuse of alcohol in Scotland (Scottish Health Survey 2014, c 2: alcohol p 38).
 The first respondent did not have to show that minimum pricing was the only conceivable option, but he had to consider a range of reasonable alternatives. This he had failed to do. The only alternative which had been considered was excise duty. The Sheffield University Group had the capability to model complex and integrated alternative measures, but they had not been commissioned to do so. Fiscal measures could be used to create an equivalent retail price below which alcohol could not be sold, thereby stopping any consumer from trading down.
 Minimum pricing had to be demonstrated as proportionate (R (Lumsdon) v Legal Services Board  3 WLR 121 at paras 24-5, 31, 33, 37, 50-5, 66-7). It had to be shown to be appropriate to achieve the objective. The first respondent had attempted to introduce a new aim, that of reducing the consumption of cheap vodka by harmful drinkers who live in poverty, with a view to reducing health inequality. That was inconsistent with the aim which has been stated to the CJEU, which was to reduce the consumption of all hazardous and harmful drinkers regardless of income. The aim of the legislation was what Parliament had understood when they passed the Bill. It was what was recorded in the BRIA. The first respondent had consistently stated that the dual aim was to reduce consumption generally, and to reduce consumption of the cheap alcohol which was favoured by hazardous and harmful drinkers (BRIA, paras 10 and 12.3). That was consistent with what Parliament had stated to the EU Commission, and what the Lord Ordinary had understood the position to be.
 There was no evidence that hazardous and harmful drinkers tended to drink alcohol that was cheap relative to its strength. The evidence relied upon in answering the questions posed by the CJEU was not before this court. It related to “heavy” drinkers, who were a different and wholly undefined class. The evidence related to studies of extreme drinkers, some of whom were consuming as much as 800 units of alcohol per week. They were not representative of the general drinking population, nor were they representative of the vast bulk of hazardous or harmful drinkers. The true position was that drinking at hazardous and harmful levels was a phenomenon which increased with affluence.
 The criticism of the statistics in the Scottish Health Survey 2014 was misplaced. The statistics took account of the difference between self-reported consumption data and sales data. It was odd that the MESAS reports since 2013 did not survey the relationship between alcohol consumption and affluence.
 The empirical evidence suggested that, whilst moderate drinkers were relatively “price elastic”, and would decrease consumption as prices increased, the same was not true of heavier drinkers. Heavy drinking by adults did not respond to higher alcohol prices. It was difficult to see why a simple price control mechanism such as minimum pricing could have a more targeted effect on hazardous and harmful drinkers, when that group was relatively price resistant.
 A number of criticisms of the Sheffield University modelling could be made. First, it did not model the correlation between affluence and hazardous and harmful drinking. It was based on a comparison of the 14% of the population who lived in poverty, as compared with the 86% of the population who did not. There was no basis upon which to suggest that the population of Scotland was too small to model on the basis of income quintiles. Secondly, the Sheffield University Report did not correspond to the stated aims of minimum pricing. It was a highly limited study, which made only limited comparisons with excise duty. It looked only at the structure of excise duty as it currently existed, without considering the possibility of other sales taxes. It focused on harmful drinkers, rather than hazardous and harmful drinkers, and on drinkers in poverty, rather than mapping across the income scale. It did not provide objective evidence, but was an exercise in advocacy.
 Minimum pricing had been proposed on the basis of an econometric model which did not investigate the links between affluence and alcohol consumption. It was predicated on a false claim that cheap alcohol was favoured by hazardous and harmful drinkers. Subsequent iterations of the Sheffield University study, which had taken into account income, revealed that alcoholic products retailing at less than 50p per unit were purchased by harmful and hazardous drinkers in the lowest two income quintiles. This meant that minimum pricing would not catch the majority of hazardous and harmful drinkers, who were found in the highest three income quintiles. Because of that, the first respondent now sought to argue that minimum pricing was designed to target harmful drinkers in poverty, rather than the broad class of hazardous and harmful drinkers across the income spectrum. Harmful drinkers in poverty made up less than 1% of the total drinking population (Sheffield University: Model-based appraisal etc (2016), Table 4.3). Minimum pricing would affect the whole drinking population, but was apparently aimed at “pricing out” only the excessive consumption of that 1%.
 The article (Meier et al: Estimated Effects of Different Alcohol Taxation and Price Policies etc (2016)) composed by the Sheffield University study authors, was to be compared with the Model-based appraisal. Both used the same methodology. The price distributions of alcohol per unit were broadly similar. Whilst sales data consistently confirmed 18-20% higher sales in Scotland, the self-reported data showed that consumption was similar in the two jurisdictions. The article’s modelling concluded that, to achieve the same 4.3% reduction in total alcohol-related deaths as 50p minimum pricing, only a 13.4% increase in existing tax rates was needed (Meier et al (supra), p 4). This was in contrast to the Model-based appraisal, which claimed that a 27% excise rate increase would be required to achieve the equivalent decrease in alcohol-related deaths (Sheffield University: Model-based appraisal etc (2016), para 4.3.1). There was no explanation as to why the appraisal did not take the same varied approach to taxation which the article’s study had. The article suggested that a minimum pricing of 50p would increase alcohol consumption amongst the more affluent hazardous and harmful drinkers, leading to a rise in their predicted mortality (Meier et al (supra), February 2016 p 15). In contrast, the Model-based appraisal demonstrated that an increase of 13.4% in duty would cause reductions in deaths across all categories of drinker and income groups. Modelling had its limitations (evidence of Dr Meier, 10 February 2010 to the Health and Sport Committee).
 On the evidence presented by the first respondent, minimum pricing was not an appropriate measure. It was not suitable to attain the objective which it pursued. Price measures were not a suitable response to the abuse of alcohol by a small minority of drinkers. The likely consequence for heavier drinkers, who were relatively price inelastic, was not that they would consume less, but that they would spend more. For those on a lower income, this would lead to a substitution of expenditure. For those who had a higher income, there would be no impact and their consumption would be unaffected by the measure.
 Minimum pricing failed the necessity test because it was not the least restrictive alternative. The Government’s travaux preparatoires for the legislation, including the BRIA, contained no discussion of the impact of the measure on intra-EU trade. The petitioners’ evidence supported the concerns of the EU Commission, specifically on the interference with intra-EU trade. The first respondent had produced no expert evidence to counter the petitioners’ evidence. Excise duty was a valuable budgetary tool, as well as a reflection of social policy. For that reason, levels of excise duty could not be harmonised across the EU. There was agreement on the way that excise duty had to be structured and applied (Directive 92/83/EEC  OJ 316/21), as well as the minimum rate of excise duty which could be charged on alcohol (Directive 92/84/EEC  OJ 316/29). Increasing excise duty on alcohol was permissible under EU law, as long as the tax levied was by reference to the quality of the finished product, and exceeded the specific minimum rates (Article 110).
 Minimum pricing would have a distortive effect on the intra-EU trade in wine, and alcohol products more generally. There were the two reports by Yarrow and Decker; the first from October 2012 and the second from May 2016. The second report took into account up-to-date statistics on alcohol pricing and price elasticity. The conclusion of both reports was that the more effective minimum pricing was, in causing consumers to switch from lower to higher priced alcohol, the bigger the distortion in intra-EU trade. Potential outcomes needed to be carefully considered and evaluated. The BRIA and the Sheffield University modelling had failed to do that. The policy effectiveness of minimum pricing was measured by a decrease in the consumption of low-cost alcohol, but it failed to take into account that consumers would switch to higher priced alcohol. The potential market impacts, and how they might be mitigated, were simply unknown. The Government had no way of knowing how effective minimum pricing could be expected to be in practice.
 A sales tax would be an alternative to tax increases. Neither tax increases, nor additional sales taxes, would disrupt the normal cost-price linkage. The prices of all alcohol-products would be increased, and suppliers would be free to reflect their own particular costs in prices at all stages of the supply chain. Duty was spread across the range of alcohol. An increase would not impede the ability of the market to reflect lower producer-costs in lower prices. A tax increase would not create the retailer-incentive problem which minimum pricing may do. Whilst an increase in tax may lead to certain products being withdrawn, that was a normal feature of a competitive market. Whilst duty increases might raise the cost of low-priced alcohol relative to higher cost drinks, the proportionality test only asked whether a measure had a less distortive effect. The obligation was to seek the least disruptive measure. It was not good enough for the first respondent to claim that the effects on the market were unpredictable.
 The question could be broken down into two stages. First, could other pricing measures achieve at least the level of desired health benefits as minimum unit pricing at 50p? Secondly, if the answer to the first question was yes, were the alternative pricing measures less distortive of the single market in alcohol products generally, and the Community-trade in wine specifically? The first respondent had to persuade the court that the answer to both of those questions was in the negative. He had failed to address either of those questions or to produce evidence which would entitle the court to answer them that way. He could not rely on practicalities, such as the suggestion that an increase in duty rates of the level required would be so unreasonable that no Parliament would consider such a measure. Excise duty could be set at “Nordic levels”.
 Even if the court did not accept the detailed criticisms of the Sheffield University Report, the study demonstrated that the first respondent’s case was untenable. There were pricing measures which could achieve at least the same anticipated health benefits as minimum pricing. A 27% excise duty rise would have the same overall impact on mortality (Sheffield University: Model-based appraisal etc (2016), Table 4.17). If the aim was to delay the deaths of harmful drinkers in poverty, this too was achievable by an increase in taxation of 70%. Imposing greater costs on moderate drinkers by increasing taxation was not a reason to prefer minimum pricing when one strand of the dual objective was to reduce the consumption of the entire population. On the face of it, pricing measures ought to be aimed at all alcohol products, cheap as well as expensive, because of the dual aim.
 The first respondent had been put on notice to produce “serious evidence” to show that increased taxation was not capable of meeting the targeted objective. No such evidence had been produced. The evidence had confirmed that increasing taxation would have the desired effect. The first respondent had failed to produce any evidence on the nature and scale of the restriction of the free movement of goods by virtue of minimum unit pricing, a comparison between the free market effects of minimum pricing and alternatives, or its effects on the proper functioning of the CMO. For that reason, he could not justify the proportionality of the measure.
 The CJEU had not said that the proportionality assessment would be the same as for alcohol in general, regardless of the issues associated with the CAP. The national court required to carry out a proportionality analysis comparing the claimed health outcomes of minimum pricing against its impact on the intra-EU trade in wine. This necessarily involved considering the objectives of the CAP and the proper functioning of the CMO. Statistics, compiled by NHS Scotland demonstrating the price distribution, and market share, of wine from different countries, were available. They had not been produced by the respondent (NHS Scotland, The price distribution of wine from different countries of origin sold in Scotland’s off-trade (August 2013)). Whilst the data did not take into account budget retailers, such as Aldi and Lidl, the data could be used to make predictions about the impact of minimum pricing on the intra-EU wine trade. Significant price rises would be required as a result of minimum unit pricing for wine produced in some EU countries (such as Spain and Italy) but not others (such as France and Germany) (Cardinal Minimum Unit Pricing Survey, Report for the Scotch Whisky Association (2016)). No evidence had been produced by the first respondent to show how the CAP/CMO currently worked, and how minimum pricing would impact on wine production.
 It is not disputed that minimum pricing has an effect equivalent to a quantitative restriction on the importation of goods. It will limit the competitiveness of alcohol which has very low production costs. A low cost producer will not be able to undercut the competition by selling below the minimum price. However, the CJEU has made it clear that minimum pricing is not rendered unlawful under Article 34 of the Treaty on the Functioning of the European Union on this basis alone. It may be justified in terms of Article 36.
THE TEST TO BE APPLIED
 The first question is whether the Lord Ordinary applied the correct test in assessing whether the legislation could be justified under Article 36 as a measure for protecting health and life.
 The CJEU stated (para 34) that, in order to be justified, the measure must be appropriate for securing the attainment of the objective pursued and must not go beyond what is necessary in order to attain it. That test is well known and derives from C-110/05 Commission v Italy  ECR I-519 (at para 59). It was applied in Sinclair Collis v Lord Advocate 2013 SC 221 (at para ). A measure is appropriate if it genuinely reflects a concern to attain the objective in a consistent and systematic manner (CJEU para 37; Sinclair Collis v Lord Advocate (supra) at para ). An Article 36 derogation cannot be sustained if life and health can be protected by a measure less restrictive of intra-EU trade, such as taxation (CJEU para 43). It is for the state to decide on the level of protection of health and life, taking into account the principle of the free movement of goods (Sinclair Collis v Lord Advocate (supra at para , citing Commission v Italy (supra) at para 65). The state requires to demonstrate proportionality; ie that the measure is necessary to achieve the declared objective and that the objective cannot be achieved by a measure less extensive or less disruptive of trade (CJEU para 53). The reasons given by the state have to be accompanied by “appropriate evidence or by an analysis of the appropriateness and proportionality” of the measure with specific evidence substantiating the state’s arguments (CJEU para 54, Sinclair Collis v Lord Advocate (supra) at para , citing Commission v Italy (supra) at para 66). The state does not require to prove that no other conceivable measure could enable the objective to be attained (CJEU para 55). The court’s task is to determine objectively:
“whether it may reasonably be concluded from the evidence submitted … that the means chosen are appropriate for the attainment of the objectives… and whether it is possible to attain those objectives by measures which are less restrictive of the free movement of goods and of the CMO” (CJEU Answer to Questions 4 and 6).
 In following Sinclair Collis v Lord Advocate (supra), and hence Commission v Italy (supra), the Lord Ordinary explained (Opinion, para ) that a measure requiring Article 36 justification had to be appropriate for securing the public interest objective and must not go beyond what was necessary to attain it. It was for the state to determine the level of protection and the way in which it was to be achieved. The state did not require to prove that no other conceivable measure could enable the objective to be attained. It did have to demonstrate that the measure was objectively proportionate; that is that it was appropriate and necessary.
 The court is unable to identify any error in the Lord Ordinary’s application of the jurisprudence of the CJEU as adopted in Sinclair Collis v Lord Advocate (supra at para ) and subsequently confirmed by the CJEU’s answers to the questions in the Reference. It follows from this that there was no error in his identification of the EU law to be applied.
THE AIM OF THE LEGISLATION
 The second question is whether the Lord Ordinary identified the correct aim of the legislation. The Lord Ordinary found (para ) that the aim was “the reduction of alcohol consumption … in particular … by hazardous and harmful drinkers”. This was principally directed towards the protection of health and life. The CJEU was satisfied, from a consideration of the Explanatory Notes and the Business and Regulatory Impact Assessment (BRIA), that the legislation did pursue the objective of protecting health and life by “reducing, in a targeted way, both the consumption of alcohol by consumers whose consumption is hazardous or harmful, and also, generally, the population’s consumption of alcohol” (para 34).
 The Lord Ordinary’s view, as derived from the material before him, is identical to that of the CJEU. It would have been surprising if any other conclusion had been reached. The aims are clearly set out in the Policy Memorandum (eg paras 3 and 16) as being to reduce alcohol consumption generally, and in particular that of harmful drinkers. That in turn is intended to reduce the impact that “alcohol misuse and overconsumption” has on “public health, crime, public services, productivity and the economy as a whole”. The idea, following the WHO (Global strategy to reduce the harmful use of alcohol (2010)), is to have a general and a targeted approach. The Lord Ordinary did not regard the consequential economic benefits as part of the aim requiring justification, although, incidentally, public order sequelae may well fall within the Article 36 protections.
APPROPRIATENESS (SUITABILITY) TO SECURE THE OBJECTIVE (AIM)
 The third question is the appropriateness of the legislation for achieving the identified objectives or aims. The petitioners maintain that there is no evidence to support the idea that a general reduction in the population’s consumption would have any significant health benefit. That is a statement which would surprise many. It runs contrary to the material, which is referred to in some detail in the Policy Memorandum (para 6 et seq), that a general increase in consumption coincides with a general increase in harm, in the form of hospital admissions and death. It effectively contradicts the WHO statement (Global Strategy (supra), para 33) that increasing the price of alcohol (ie in general) is one of the most effective ways of reducing its harmful use.
 A general increase in consumption should not be understood as meaning that everyone in Scotland has increased his or her consumption in a uniform manner. The coincidence between an increase in general and the levels of harm indicates that some of the population have increased their consumption at some point into the harmful or hazardous categories. The reason for that appears to be, at least in large part, the availability of cheap alcohol or, put another way, the increasing affordability of that category of alcohol through the off-trade. The fact that the legislation would affect moderate drinkers in some way does not detract from the legitimacy of the aim as a measure designed for the general protection of public health and life.
 Whether there is what might be described as controversy or not, there was ample material from which the Government could conclude, as they did, that there is a direct link between the price of alcohol and its rate of consumption in the general population. There was sufficient information from which the Government could hold that there is an equally clear link between the consumption of alcohol and health problems. This is so, even if the person who is constantly a moderate drinker may be fortunate enough not to succumb to any of these problems over the course of time.
 As the CJEU said (para 36), following a similar statement from the Advocate General (para 127), in determining that it was satisfied that the legislation was appropriate:
“it does not seem unreasonable to consider that a measure that sets a minimum selling price …, the very specific aim of which is to increase the price of cheap alcoholic drinks, is capable of reducing the consumption of alcohol, in general, and the hazardous or harmful consumption of alcohol, in particular, given that drinkers whose consumption can be so described purchase, to a great extent, cheap alcoholic drinks.”
Since the legislation forms part of a wider political strategy to combat the effects of alcohol, the CJEU was also content that it met the criterion of attaining the objective in a consistent and systematic manner and was thus appropriate (para 39).
 It is important to observe that it is for the state to decide upon the degree of protection. It is afforded by both the CJEU and this court a margin of discretion. Before looking, therefore, at what appears to be the core issue of whether life and health can as effectively be protected by measures less restrictive of trade and competition, notably tax increases, it is worth pausing to reflect on one central feature; that is the nature and extent of the problem in this jurisdiction.
 The societal, family and personal effects of excessive alcohol consumption in Scotland are difficult to over-estimate. In some comedic settings they form an unfortunate, if distorted, caricature of the Scottish character. The effect of excessive consumption on the nation’s health, levels of crime and productivity is notorious and hardly needs exposition, since they are apparent in daily life, especially to those practicing in the courts. According to the Government, the annual cost of excessive alcohol consumption can be estimated in billions of pounds. It can come as no surprise therefore that the Government, in accordance with the WHO Global Strategy (supra), has attempted to identify the extent of this particular, pervasive societal concern and determined to devise measures which will reduce its effects to an acceptable minimum. A range of measures, which have already been taken, were detailed in the BRIA under reference to the strategy document (Changing Scotland’s Relationship with Alcohol: A framework for Action (2009)). The stage was reached, however, when the Government recognised that the “Key component missing … has been an intervention to address the low price of alcohol” (BRIA, summary para 6).
 In the Policy Memorandum, the Government set out source figures to illustrate what was perceived to be the problem and its source. The problem was well illustrated by the stark fact that enough alcohol is being sold to enable each and every adult to exceed the recommended male weekly guideline maximum of 21 units on each and every week of the year. This is driven by the off-trade (supermarkets) as the primary source, given that the average price of a unit of alcohol there is 45p, as distinct from £1.34 in the on-trade (pubs). The excessive consumption of alcohol is causing a significant increase in hospital admissions. Of considerable importance is the identification of an acute problem amongst the less affluent. The lowest income quintile’s alcohol related hospital discharges are running at a rate some 7.5 times greater than those in the highest income quintile. Alcohol mortality is 6 times higher. This, by any gauge, is a major concern to any responsible democratic Government (see WHO’s Global Strategy (supra) especially targeted groups at c 1, para 11). There has been no alteration in the group targeted by the Government. The particular impact on deprived communities was identified at any early stage. It is referred to, for example, in the Policy Memorandum (eg paras 10 and 16).
 Although a raft of statistical material was produced, there are a number of straightforward propositions which might be regarded as self-evident, although they are also backed by many of the papers produced and, in any event, largely accepted by the CJEU. The level of consumption of alcohol is affected by pricing. Higher prices will have a greater influence on those with less money; notably the lowest earners including, to a degree, the young. Those wishing to consume large quantities of alcohol, whether rich or poor, will obtain most of it from the off-trade.
 The Policy Memorandum was supported by the BRIA which set out, in relatively clear terms, the Government’s view based on their interpretation of the available material. Both of these documents paint a picture which illustrates how minimum pricing will target cheap alcohol, which again is hardly surprising since it will have little or no effect at all on higher priced products. The critical feature, however, is that it will target harmful and hazardous drinkers, since they are the group who tend to buy cheap alcohol (see Lord Ordinary’s Opinion, para ). It will have a significant impact on those in the lower quintile of affluence since, unlike richer citizens, they will not be able to trade down to products which are cheaper than those which they are presently consuming. It is this poorer section of society which is most affected by cheap alcohol in terms of hospital admissions and mortality. This may not be surprising given economic differentials and, in particular, the comparative levels of housing quality and diet. The fact that minimum pricing may not, to the same extent, affect those who are more affluent, is of peripheral significance. These richer persons tend not to suffer to the same extent as harmful and hazardous drinkers in the lower quintile of affluence, whose health and life is at greatest risk. As the Lord Ordinary held (Opinion, para ) there is “clear evidence that the greatest alcohol-related harm is experienced by those who live in the most deprived areas”. The statistics on hospital discharges and mortality are striking examples of the differentials involved.
 It is possible to deploy statistics and related arguments in an attempt to rebut the figures founded upon in support of the Bill and to counter the conclusions drawn by the Government in the Policy Memorandum, the Explanatory Notes and the BRIA. The fundamental proposition remains, however, that there was and is ample objective material to support the proposition, as the Government have concluded, that a two pronged approach, of the type described, is the best way forward. Some of the counter arguments may have some merit, at least in the context of a courtroom debate. They do not detract from the central point that such material as the Government have produced amply justifies their view, notwithstanding those of others to the contrary.
 As the Lord Ordinary held (para ), the major problem is the excessive consumption of cheap alcohol. In that respect, he did not fail to give appropriate weight to the lesser effect on the more affluent harmful and hazardous drinkers. That group may well be affected less, although there is material justifying a conclusion that they are still affected to a degree. The Lord Ordinary cannot be faulted in finding that there was evidence from which it could be inferred that minimum pricing was an appropriate method of securing the objective by tackling the specific consumption of cheap alcohol. The contention that the Lord Ordinary failed to “exercise the due degree of critical scrutiny” of the evidence is rejected. In that respect the issue is not what the Lord Ordinary may have concluded as a matter of fact following upon such a scrutiny, but whether he was entitled to find that objectively it could reasonably be concluded from the material submitted by the Government that the measure was appropriate to attain the objective (CJEU, Answer to questions 4 and 6). Since the test is objective, the question of whether the material was put before the Parliament during the process of enactment is irrelevant.
PROPORTIONALITY: NECESSITY AND THE LESS RESTRICTIVE ALTERNATIVE
 The fourth question relates to the necessity of the legislation and whether a less restrictive measure, notably an increase in tax might achieve the same or a similar objective. In this area, as the first respondent accepts, the Government cannot pray in aid the objective of a general reduction in alcohol consumption as justifying the legislation in terms of Article 36, since an increase in tax across the board could achieve that simple objective, assuming that the increase was reflected in retail prices. It is important to recognise, as the Lord Ordinary did (Opinion, para ), that the object of the Government’s policy and the legislation is not to eradicate alcohol consumption. In that sense the position is quite different from tobacco.
 For many, the consumption of alcohol is a great social lubricant. Its use in a responsible manner is an important and positive feature in Scottish culture. The general intent, again as the Lord Ordinary noted (Opinion, para ), is to “get people to build a healthy and sensible relationship with alcohol”; not to stop them drinking altogether or even to drink moderately at all times. This general intent may itself prove beneficial on a number of fronts. The targeted objective nevertheless remains firmly the consumption of cheap alcohol by those whose health is most likely to be adversely affected by it.
 In practical terms, the measure achieves the targeted objective by setting a floor price below which alcohol cannot be sold. Alcohol will not be sold for less than 50p per unit. Those who currently consume cheap alcohol at a harmful and hazardous level will not be able to switch to another product to maintain their consumption levels. The true area for debate is whether modification of taxation, within the permissible bounds of EU law, can achieve similar results in targeting cheap alcohol as is undoubtedly achievable with minimum pricing. This is also what the Lord Ordinary considered to be a critical issue; “the comparative effectiveness of minimum pricing and the alternative measures” (Opinion, para ). The test is described by the CJEU (para 41) as one of whether life and health can be “as effectively protected” by tax changes. That is precisely the test adopted by the Lord Ordinary (Opinion, paras  and ). If the alternative would not be “as effective”, minimum pricing would be necessary and proportionate. He did not say that the alternative had to have precisely the same effect.
 Before diving deeper into this area, a few further principles require to be highlighted, in the context of the state’s area of responsibility and that of the national courts. The first, at the risk of unnecessary repetition, is that it is primarily for the state to determine the level of protection which it wishes to afford to its citizens’ life and health (CJEU, para 52) and the means by which that level can be achieved. The state has a margin of appreciation in that regard.
 The level of scrutiny which the court should apply, when determining proportionality, is not, as the petitioners would have it, “high”; nor is it “low”, as was suggested in the submissions in Sinclair Collis v Lord Advocate (supra, at para ). It may be that, as the Advocate General subtly put it, it should simply be “marked by a certain degree of restraint” (para 82) by the courts recognizing the distinction between what may be a political decision for Government and a decision for the courts concerning the legality of the measure in terms of EU law. Whilst the state requires to demonstrate (at least at first instance) the proportionality of the measure by reference to appropriate “evidence” or an “analysis of the appropriateness and proportionality of the” measure (CJEU para 54), it does not have to prove that no other conceivable measure could enable the legitimate objective to be attained (CJEU, para 55). Nevertheless, the court requires to examine objectively whether the aims can be attained by a measure less restrictive of intra-EU trade.
 The court requires to carry out a structured proportionality exercise. The task is not simply to balance the benefits of a measure such as minimum pricing with the impact on intra-EU trade. That is a matter for the democratically elected Government, and not the court. It is not an attempt to balance the cost-benefits of minimum pricing with taxation measures. This too would put the court in the position of substituting its own view of where the balance ought to be struck between protecting health and interfering with the market.
 The CJEU has set out in clear terms how the second stage of the structured proportionality assessment is to be carried out. The court has to compare the effectiveness of minimum unit pricing in achieving the targeted objective, with other measures which could also achieve that objective, and which would be less restrictive of intra-EU trade. Only if minimum pricing is the more effective way of achieving the objective will the interference with the market be necessary and therefore proportionate under Article 36. The identification and selection of other measures to compare with minimum pricing, such as increased tax, assumes that the other measures will be less restrictive of intra-EU trade, and interfere less with the market.
 The fact that the precise effect of the relevant measure cannot be known is a relevant factor when determining whether what is an experimental piece of legislation meets the requirements of proportionality (CJEU, para 57). The certainty of the provision and the ease with which it can be understood, managed and supervised, when compared with the alternative, is also a factor in this equation (Sinclair Collis v Lord Advocate (supra), at para  following Commission v Italy (supra) at para 67).
 In all of this, the starting point must be that it has been demonstrated, for reasons already given, that minimum pricing would achieve the legitimate Article 36 objective of the protection of life and health. The issue then returns to whether raising the level of tax on alcohol would achieve the same or similar objective (“as effective”) with a less restrictive effect. Here, of course, the elephant in the room is the fact that the Scottish Government has no power to raise taxation on alcohol. That is a matter reserved to the United Kingdom Government. Conversely, the UK Government has little responsibility for the health of the inhabitants of Scotland; that being a devolved matter. It was agreed by the parties that, in the EU context, this division of power and responsibility is irrelevant because the member state is the UK. It required to organize itself accordingly. That may be so, but it does produce a curious anomaly in the context of a legal argument that increasing tax is a viable alternative, when the political reality is that it is clearly not.
 Although the Lord Ordinary is said by the petitioners to have dismissed the alternative of tax increases “peremptorily”, it was this very issue that he expressly regarded as “critical” and which he spent a considerable part of his Opinion (paras  to ) analysing; notably the central contention for the first respondent that the alternative of tax was “more complicated and less predictable” (para ) than the petitioners maintained. The analysis was conducted within a framework which accepted the petitioners’ argument about the adverse consequences for the free movement of goods. The Lord Ordinary accepted (para ) the basic tenet that if the tax alternatives were as effective, the measure would fall foul of Article 34 as they would be less of an obstacle to that movement. Ultimately, as already described, he held that there was objective justification for the view that increasing tax would not be “as effective” for a number of reasons. It is difficult to detect any fault in that reasoning.
 The faults alleged were said to include failing to have proper regard to whether minimum pricing was needed at all in light of other measures said to be achieving benefits in relation to a reduction in overall consumption. Once it is accepted that minimum pricing will achieve the objective, notably in relation to the targeted sections of society, in the context of the other strategic measures operating, the comparative success of the other measures in relation to the general populace or the targeted parts is little to the point. The material supporting the Bill explained the lacuna, which existed in relation to cheap alcohol and which required filling (eg Explanatory Notes, para 6).
 There is a danger, in what follows, that the court is engaging in a process of re-analysis which it has already cautioned against. However, in exploring whether the Lord Ordinary’s assessment is flawed, some comparative analysis may be justified in order to reach a view. Some further criticisms of the Lord Ordinary, in the grounds of appeal, do require express consideration. First, the Lord Ordinary did not fail to give the Opinion of the EU Commission due weight. He specifically referred to it (Opinion, para ) and described its content in some detail. He stated that it was of interest, and so it was. He correctly stated that it did not bind him. Thus, he is explaining that he required to use his own judgment in determining the issue. The Lord Ordinary did have proper regard to the tobacco cases (C-197/08 Commission v France  ECR I-1599; C-198/08 Commission v Austria  ECR I-1645; C-221/08 Commission v Ireland  ECR I-1669). He referred to them both in the context of noting the petitioners’ submissions (Opinion, para ) and when analyzing their significance (ibid para ). He did not specifically mention the older C-216/98 Commission v Greece  ECR I-8921, but he would have been aware of its ratio, from a consideration of the more recent decisions, as being that Article 36 did not provide a defence to Article 9 of the Council Directive (95/59/EC) on the harmonization of tax on tobacco.
 The fundamental problem with an increase in tax is simply that it does not produce a minimum price. As stated in the Memorandum, many supermarkets have, in the past, sold alcohol at below cost. They have absorbed any tax increases by off-setting them against the price of other products unrelated to alcohol. Cheap alcohol is perceived as a draw, lure or enticement to pull shoppers into the particular retailer’s premises and away from those of the competition.
 The manner in which the supermarkets operate was described by Prof Ludbrook and Ms Gillan in their evidence to the Parliamentary Committee. It may well be, stated broadly, that, in the case of tobacco, “increases in excise duties must sooner or later be reflected in increased retail selling prices” (CJEU, para 44, citing C-216/98 Commission v Greece (supra), at para 31 and C-197/08 Commission v France (supra), at para 52). However, especially in the case of alcohol, they do not have to be, and very often are not, reflected in the retail price of the particular product or class of product, viz. alcohol. The scope, as with increased tax, for variation, manipulation and cross-subsidisation (absorption) far exceeds that which is possible in a minimum pricing context. Of course, it would be possible to impose a ban on sales at below the cost levels of each product and then levy increased tax upon that. Such a model would be complex and difficult to enforce, when set against one of straightforward minimum pricing. For example, there would require to be an accurate assessment of what the cost might have been.
 The advantage of the proposed minimum pricing system, so far as protecting health and life is concerned, is that it is linked to the strength of the alcohol. Current EU tax arrangements relate to different types of product (wine, spirits, beer and cider etc) each of which has a range of ABVs. As the Lord Ordinary discovered (Opinion, para ), they do not permit taxation of wines at variable rates according to the strength of alcoholic content. Minimum pricing permits lower pricing to be charged for lower strength alcohol.
 Minimum pricing is targeted at harmful and hazardous drinkers because it is those groups that tend to purchase cheap alcohol; not moderate drinkers who select from higher quality products. In this area, the Lord Ordinary (para ) was careful to reference the BRIA which stated (para 4.3) that there was evidence that across the board tax increases did not have a targeted effect on those at risk of harm. He cited the reference for that (Gruenwald et al: Alcohol Prices, Beverage Quality and the Demand for Alcohol etc (2006)). He explained the reason for that, as given in the BRIA, as being that harmful and hazardous drinkers consumed a disproportionate amount of cheaper products. He cited the reference for that too, as emanating from Booth et al (Independent Review of the Effects of Alcohol Pricing and Promotion Part A: Systematic Reviews (2008). The footnote in the BRIA quoted the Gruenwald paper as explaining that price increases targeted at the lowest cost brands would produce a greater reduction in sales than across the board price increases. This is because trading down is not possible where the alcohol is at its minimum price. This phenomenon was spoken to by several of the witnesses who gave evidence to the Parliamentary Committee. This targeting cannot be achieved by a tax increase directed at low cost products, because EU law will not permit there to be anything other than a uniform rate.
 Once this is understood, it can, as the Lord Ordinary concluded, be seen that whatever arguments may be deployed against it, there is material (“evidence”) which demonstrates that the alternative of increased tax, with or without a prohibition on below cost sales, is less effective than minimum pricing. The Government are entitled to found upon this as grounding a valid Article 36 justification. In all of this the Lord Ordinary’s reasoning is sound. Furthermore, assuming that any practical tax increase within the EU setting would involve across the board increases, albeit perhaps on different types of product, such increases would have a disproportionate, undesirable and unnecessary effect on moderate drinkers, who do not generally represent a significant problem in societal terms, at least of the type requiring to be addressed.
 It is important to return to recognise, as the CJEU did (para 57), the provisional or experimental nature of the legislation. What is unknown is the reaction of the off-trade to a minimum pricing regime. The manner in which the supermarkets might behave is necessarily unclear because it has not been disclosed, assuming that it has even been formulated in any concrete manner. Little or no information is available in this area. That is not because the Government or others have failed to make the appropriate inquiries, but because of commercial sensitivities.
 The only way in which minimum pricing can be tested is by trialing it; which is what the Government seek to do. The petitioners complain that the Government ought to have produced models on the effect of the measure on intra-EU trade. However, that is largely impossible without a clear understanding of the commercial thinking of those in control of the retail market (the supermarkets). That thinking is unlikely to be revealed to public authorities (see eg Dr Rice’s evidence to the Parliamentary Committee; BRIA paras 5.114-118). There are an almost infinite number of variables, depending upon: what margins are applied to products formerly selling below the minimum price; what price (above the minimum) might be achieved for those products; whether those already above the minimum will be priced at a level which maintains the differential between them; and what the public may still regard as “cheap alcohol”. It is partly because of all this that the experimental nature of the legislation becomes a consideration of some significance.
 What is known is that the legislation is likely to have an effect on EU trade with the UK, albeit that Scotland is a relatively small part of the UK market, even in alcohol. In EU market terms the effect might be described as relatively minor. The on-trade is unlikely to be adversely affected at all. No doubt some wine from Bulgaria, Romania and Portugal may lose a competitive edge. Their share of the market too is very small, but there will be an effect on the competitive nature of some wines and beers from other EU states. Cheap French brandy may be affected, even if, so far as spirits are concerned, the greater impact will be on domestically produced vodka, whisky and cider. It is hardly necessary, and probably impracticable, to offer any more precise prediction as to the general effects. They are recognised, but the task for the Government was, and is, to take the adverse effect into account in deciding whether the health and life benefits outweigh them. That is what they have done.
 Proportionality must involve a balance. The selection by the Government of the MUP in the legislation involves such a balance, not only in relation to trade with the EU, but having regard to the effect on domestic producers and suppliers. The documents accompanying the Bill, notably the BRIA, contain a detailed analysis of such matters. No doubt they were considered when the figure of 50p per unit was selected. Such a figure, on the material produced, will reduce consumption amongst harmful and hazardous drinkers in that quintile of the population whose health is affected most by the consumption of cheap alcohol. The benefits of this are well documented.
 In all the circumstances, and as a matter of objective fact, to be assessed by the national court in accord with the task set by the CJEU, it is reasonable to conclude that alternative measures, including increases in taxation, are not capable of protecting life and health as effectively as minimum pricing, while being less restrictive of trade. In so finding, it is of course almost inevitable that there will be evidence pointing to a different conclusion. However, as the CJEU has ruled, the state does not have to demonstrate that no other conceivable measure could enable the objective to be attained.
THE COMMON ORGANISATION OF AGRICULTURAL MARKETS REGULATION (EU No 1308/2013)
 The CJEU determined (Answer to Question 1) that the Regulation EU No 1308/2013 did not preclude a measure introducing minimum pricing for wine, provided that the measure was, as a matter of fact, an appropriate means of securing the objective of the protection of life and health and, taking into account the objectives of the Common Agricultural Policy and the proper functioning of the Common Organisation of the agricultural Market, did not go beyond what was necessary to attain that objective. The CJEU advised (para 40) that an analysis of proportionality, separate from that carried out for the purposes of Article 36, did not have to be carried out. In these circumstances, although there may be elements of the Lord Ordinary’s Opinion that were not quite accurate under this chapter, his fundamental reasoning, that the lack of an equivalent to Article 36 in the Regulation did not support the conclusion that the Regulation was exhaustive of the issue, was correct. In this case the issue in relation to wine under the Regulation raises points which are almost identical to those relative to alcohol generally under Article 36. Following the advice of the CJEU, no separate analysis is required in this case. The result is the same. Minimum pricing would attain the stated objective of protecting life and health, particularly of harmful and hazardous drinkers in the lower quintile, gauged by wealth, of the population. That may be a relatively small percentage of the population but it is significant in numerical terms. The issue then becomes one of whether it has been demonstrated that taxation, whether or not combined with other measures, would not be as effective in attaining the objective. For the reasons already provided in relation to Article 36, there is ample material from the Government (albeit contradicted by some documents produced by the petitioners) from which it can be objectively concluded that taxation, on its own or in combination, would not achieve the same or a similar objective. The argument under the Regulation fails for the same reasons as those relative to the Treaty itself.
 For these reasons, the grounds submitted in the reclaiming motion are not well founded. The Lord Ordinary directed himself correctly on European Union law. He applied that law accurately to the facts which he found demonstrated by the material before him. It follows, from what has already been described, that the considerable amount of material produced after his decision and founded upon by both parties, could not have had a material bearing on the Lord Ordinary’s reasoning; quite the contrary. In these circumstances, the reclaiming motion is refused and the court will adhere to the interlocutor of 3 May 2013 refusing the petition for judicial review of the measure.